{
  "id": "veil-persian-tokenizer-efficiency",
  "title": "Persian tokenizer efficiency across models",
  "version": "1.0.0",
  "measuredAt": "2026-08-19",
  "license": "CC-BY-4.0",
  "measures": "Tokens spent on the same meaning in Persian versus English, per model tokenizer. This is a cost and context-window measurement. It is not a measure of model quality, Persian fluency, or reasoning.",
  "method": "Every passage in the paired corpus is encoded with the model's own tokenizer file, read from HuggingFace at the recorded commit. Counts are exact, not estimated. A model whose encode/decode round-trip does not reproduce the input is dropped rather than reported.",
  "corpus": {
    "id": "veil-persian-tokenizer-corpus",
    "version": "1.0.0",
    "items": 28,
    "faCharacters": 3065,
    "enCharacters": 3592
  },
  "harness": "scripts/tokenizer-bench.mjs",
  "verification": {
    "roundTrip": "every published row reproduces all 56 passages exactly",
    "reference": "cross-checked against huggingface/tokenizers via scripts/verify-tokenizer.py; see verifiedAgainstReference"
  },
  "models": [
    {
      "id": "persianmind",
      "label": "PersianMind",
      "vendor": "University of Tehran",
      "repo": "universitytehran/PersianMind-v1.0",
      "mirror": false,
      "persianSpecific": true,
      "note": "Llama-2 with a Persian-extended vocabulary. On this board to answer the obvious question: does extending the vocabulary for Persian actually pay?",
      "revision": "af603eeb074138e2a613fbc95d89f018afbd3041",
      "fileSha256": "add5ca15f72824c65d84025377ab32a53096095bafe8dcd55ce0871b8af1f91e",
      "measuredAt": "2026-08-19",
      "vocabSize": 41510,
      "fingerprint": "5d3f077a633414ae",
      "items": [
        {
          "id": "news-01",
          "register": "news",
          "faTokens": 35,
          "enTokens": 34,
          "ratio": 1.029
        },
        {
          "id": "news-02",
          "register": "news",
          "faTokens": 36,
          "enTokens": 38,
          "ratio": 0.947
        },
        {
          "id": "news-03",
          "register": "news",
          "faTokens": 33,
          "enTokens": 27,
          "ratio": 1.222
        },
        {
          "id": "technical-01",
          "register": "technical",
          "faTokens": 35,
          "enTokens": 29,
          "ratio": 1.207
        },
        {
          "id": "technical-02",
          "register": "technical",
          "faTokens": 32,
          "enTokens": 24,
          "ratio": 1.333
        },
        {
          "id": "technical-03",
          "register": "technical",
          "faTokens": 31,
          "enTokens": 24,
          "ratio": 1.292
        },
        {
          "id": "technical-04",
          "register": "technical",
          "faTokens": 32,
          "enTokens": 27,
          "ratio": 1.185
        },
        {
          "id": "policy-01",
          "register": "policy",
          "faTokens": 36,
          "enTokens": 27,
          "ratio": 1.333
        },
        {
          "id": "policy-02",
          "register": "policy",
          "faTokens": 33,
          "enTokens": 22,
          "ratio": 1.5
        },
        {
          "id": "conversational-01",
          "register": "conversational",
          "faTokens": 26,
          "enTokens": 27,
          "ratio": 0.963
        },
        {
          "id": "conversational-02",
          "register": "conversational",
          "faTokens": 31,
          "enTokens": 24,
          "ratio": 1.292
        },
        {
          "id": "conversational-03",
          "register": "conversational",
          "faTokens": 27,
          "enTokens": 26,
          "ratio": 1.038
        },
        {
          "id": "scientific-01",
          "register": "scientific",
          "faTokens": 40,
          "enTokens": 31,
          "ratio": 1.29
        },
        {
          "id": "scientific-02",
          "register": "scientific",
          "faTokens": 25,
          "enTokens": 28,
          "ratio": 0.893
        },
        {
          "id": "scientific-03",
          "register": "scientific",
          "faTokens": 35,
          "enTokens": 26,
          "ratio": 1.346
        },
        {
          "id": "product-01",
          "register": "product",
          "faTokens": 24,
          "enTokens": 25,
          "ratio": 0.96
        },
        {
          "id": "product-02",
          "register": "product",
          "faTokens": 30,
          "enTokens": 24,
          "ratio": 1.25
        },
        {
          "id": "finance-01",
          "register": "finance",
          "faTokens": 28,
          "enTokens": 26,
          "ratio": 1.077
        },
        {
          "id": "finance-02",
          "register": "finance",
          "faTokens": 23,
          "enTokens": 22,
          "ratio": 1.045
        },
        {
          "id": "education-01",
          "register": "education",
          "faTokens": 31,
          "enTokens": 27,
          "ratio": 1.148
        },
        {
          "id": "education-02",
          "register": "education",
          "faTokens": 18,
          "enTokens": 25,
          "ratio": 0.72
        },
        {
          "id": "legal-01",
          "register": "legal",
          "faTokens": 29,
          "enTokens": 34,
          "ratio": 0.853
        },
        {
          "id": "legal-02",
          "register": "legal",
          "faTokens": 21,
          "enTokens": 22,
          "ratio": 0.955
        },
        {
          "id": "literary-01",
          "register": "literary",
          "faTokens": 34,
          "enTokens": 26,
          "ratio": 1.308
        },
        {
          "id": "literary-02",
          "register": "literary",
          "faTokens": 31,
          "enTokens": 20,
          "ratio": 1.55
        },
        {
          "id": "medical-01",
          "register": "medical",
          "faTokens": 23,
          "enTokens": 22,
          "ratio": 1.045
        },
        {
          "id": "instruction-01",
          "register": "instruction",
          "faTokens": 32,
          "enTokens": 23,
          "ratio": 1.391
        },
        {
          "id": "instruction-02",
          "register": "instruction",
          "faTokens": 28,
          "enTokens": 27,
          "ratio": 1.037
        }
      ],
      "faTokens": 839,
      "enTokens": 737,
      "ratio": 1.138,
      "faTokensPerChar": 0.2737,
      "enTokensPerChar": 0.2052,
      "faCharsPerToken": 3.653,
      "rank": 1,
      "tied": false
    },
    {
      "id": "gemma",
      "label": "Gemma",
      "vendor": "Google",
      "repo": "Xenova/gemma-tokenizer",
      "mirror": true,
      "persianSpecific": false,
      "note": "256k vocabulary — the largest on this board.",
      "revision": "fbc8c39d11899b177f1fe8f1d25bd25fbee66fad",
      "fileSha256": "05e97791a5e007260de1db7e1692e53150e08cea481e2bf25435553380c147ee",
      "measuredAt": "2026-08-19",
      "vocabSize": 256000,
      "fingerprint": "36f857605252c885",
      "items": [
        {
          "id": "news-01",
          "register": "news",
          "faTokens": 38,
          "enTokens": 34,
          "ratio": 1.118
        },
        {
          "id": "news-02",
          "register": "news",
          "faTokens": 36,
          "enTokens": 35,
          "ratio": 1.029
        },
        {
          "id": "news-03",
          "register": "news",
          "faTokens": 37,
          "enTokens": 24,
          "ratio": 1.542
        },
        {
          "id": "technical-01",
          "register": "technical",
          "faTokens": 37,
          "enTokens": 28,
          "ratio": 1.321
        },
        {
          "id": "technical-02",
          "register": "technical",
          "faTokens": 34,
          "enTokens": 22,
          "ratio": 1.545
        },
        {
          "id": "technical-03",
          "register": "technical",
          "faTokens": 34,
          "enTokens": 24,
          "ratio": 1.417
        },
        {
          "id": "technical-04",
          "register": "technical",
          "faTokens": 33,
          "enTokens": 26,
          "ratio": 1.269
        },
        {
          "id": "policy-01",
          "register": "policy",
          "faTokens": 37,
          "enTokens": 23,
          "ratio": 1.609
        },
        {
          "id": "policy-02",
          "register": "policy",
          "faTokens": 34,
          "enTokens": 18,
          "ratio": 1.889
        },
        {
          "id": "conversational-01",
          "register": "conversational",
          "faTokens": 29,
          "enTokens": 25,
          "ratio": 1.16
        },
        {
          "id": "conversational-02",
          "register": "conversational",
          "faTokens": 29,
          "enTokens": 22,
          "ratio": 1.318
        },
        {
          "id": "conversational-03",
          "register": "conversational",
          "faTokens": 27,
          "enTokens": 25,
          "ratio": 1.08
        },
        {
          "id": "scientific-01",
          "register": "scientific",
          "faTokens": 42,
          "enTokens": 30,
          "ratio": 1.4
        },
        {
          "id": "scientific-02",
          "register": "scientific",
          "faTokens": 34,
          "enTokens": 26,
          "ratio": 1.308
        },
        {
          "id": "scientific-03",
          "register": "scientific",
          "faTokens": 35,
          "enTokens": 25,
          "ratio": 1.4
        },
        {
          "id": "product-01",
          "register": "product",
          "faTokens": 34,
          "enTokens": 24,
          "ratio": 1.417
        },
        {
          "id": "product-02",
          "register": "product",
          "faTokens": 34,
          "enTokens": 20,
          "ratio": 1.7
        },
        {
          "id": "finance-01",
          "register": "finance",
          "faTokens": 30,
          "enTokens": 24,
          "ratio": 1.25
        },
        {
          "id": "finance-02",
          "register": "finance",
          "faTokens": 29,
          "enTokens": 20,
          "ratio": 1.45
        },
        {
          "id": "education-01",
          "register": "education",
          "faTokens": 35,
          "enTokens": 26,
          "ratio": 1.346
        },
        {
          "id": "education-02",
          "register": "education",
          "faTokens": 23,
          "enTokens": 24,
          "ratio": 0.958
        },
        {
          "id": "legal-01",
          "register": "legal",
          "faTokens": 31,
          "enTokens": 32,
          "ratio": 0.969
        },
        {
          "id": "legal-02",
          "register": "legal",
          "faTokens": 25,
          "enTokens": 21,
          "ratio": 1.19
        },
        {
          "id": "literary-01",
          "register": "literary",
          "faTokens": 33,
          "enTokens": 26,
          "ratio": 1.269
        },
        {
          "id": "literary-02",
          "register": "literary",
          "faTokens": 29,
          "enTokens": 20,
          "ratio": 1.45
        },
        {
          "id": "medical-01",
          "register": "medical",
          "faTokens": 31,
          "enTokens": 19,
          "ratio": 1.632
        },
        {
          "id": "instruction-01",
          "register": "instruction",
          "faTokens": 29,
          "enTokens": 23,
          "ratio": 1.261
        },
        {
          "id": "instruction-02",
          "register": "instruction",
          "faTokens": 29,
          "enTokens": 26,
          "ratio": 1.115
        }
      ],
      "faTokens": 908,
      "enTokens": 692,
      "ratio": 1.312,
      "faTokensPerChar": 0.2962,
      "enTokensPerChar": 0.1927,
      "faCharsPerToken": 3.376,
      "rank": 2,
      "tied": true
    },
    {
      "id": "gemma-2",
      "label": "Gemma 2",
      "vendor": "Google",
      "repo": "Xenova/gemma2-tokenizer",
      "mirror": true,
      "persianSpecific": false,
      "note": "",
      "revision": "550a07caea23182155d4803a36a81c6f33040024",
      "fileSha256": "3f289bc05132635a8bc7aca7aa21255efd5e18f3710f43e3cdb96bcd41be4922",
      "measuredAt": "2026-08-19",
      "vocabSize": 256000,
      "fingerprint": "5267192a42d67ba7",
      "items": [
        {
          "id": "news-01",
          "register": "news",
          "faTokens": 38,
          "enTokens": 34,
          "ratio": 1.118
        },
        {
          "id": "news-02",
          "register": "news",
          "faTokens": 36,
          "enTokens": 35,
          "ratio": 1.029
        },
        {
          "id": "news-03",
          "register": "news",
          "faTokens": 37,
          "enTokens": 24,
          "ratio": 1.542
        },
        {
          "id": "technical-01",
          "register": "technical",
          "faTokens": 37,
          "enTokens": 28,
          "ratio": 1.321
        },
        {
          "id": "technical-02",
          "register": "technical",
          "faTokens": 34,
          "enTokens": 22,
          "ratio": 1.545
        },
        {
          "id": "technical-03",
          "register": "technical",
          "faTokens": 34,
          "enTokens": 24,
          "ratio": 1.417
        },
        {
          "id": "technical-04",
          "register": "technical",
          "faTokens": 33,
          "enTokens": 26,
          "ratio": 1.269
        },
        {
          "id": "policy-01",
          "register": "policy",
          "faTokens": 37,
          "enTokens": 23,
          "ratio": 1.609
        },
        {
          "id": "policy-02",
          "register": "policy",
          "faTokens": 34,
          "enTokens": 18,
          "ratio": 1.889
        },
        {
          "id": "conversational-01",
          "register": "conversational",
          "faTokens": 29,
          "enTokens": 25,
          "ratio": 1.16
        },
        {
          "id": "conversational-02",
          "register": "conversational",
          "faTokens": 29,
          "enTokens": 22,
          "ratio": 1.318
        },
        {
          "id": "conversational-03",
          "register": "conversational",
          "faTokens": 27,
          "enTokens": 25,
          "ratio": 1.08
        },
        {
          "id": "scientific-01",
          "register": "scientific",
          "faTokens": 42,
          "enTokens": 30,
          "ratio": 1.4
        },
        {
          "id": "scientific-02",
          "register": "scientific",
          "faTokens": 34,
          "enTokens": 26,
          "ratio": 1.308
        },
        {
          "id": "scientific-03",
          "register": "scientific",
          "faTokens": 35,
          "enTokens": 25,
          "ratio": 1.4
        },
        {
          "id": "product-01",
          "register": "product",
          "faTokens": 34,
          "enTokens": 24,
          "ratio": 1.417
        },
        {
          "id": "product-02",
          "register": "product",
          "faTokens": 34,
          "enTokens": 20,
          "ratio": 1.7
        },
        {
          "id": "finance-01",
          "register": "finance",
          "faTokens": 30,
          "enTokens": 24,
          "ratio": 1.25
        },
        {
          "id": "finance-02",
          "register": "finance",
          "faTokens": 29,
          "enTokens": 20,
          "ratio": 1.45
        },
        {
          "id": "education-01",
          "register": "education",
          "faTokens": 35,
          "enTokens": 26,
          "ratio": 1.346
        },
        {
          "id": "education-02",
          "register": "education",
          "faTokens": 23,
          "enTokens": 24,
          "ratio": 0.958
        },
        {
          "id": "legal-01",
          "register": "legal",
          "faTokens": 31,
          "enTokens": 32,
          "ratio": 0.969
        },
        {
          "id": "legal-02",
          "register": "legal",
          "faTokens": 25,
          "enTokens": 21,
          "ratio": 1.19
        },
        {
          "id": "literary-01",
          "register": "literary",
          "faTokens": 33,
          "enTokens": 26,
          "ratio": 1.269
        },
        {
          "id": "literary-02",
          "register": "literary",
          "faTokens": 29,
          "enTokens": 20,
          "ratio": 1.45
        },
        {
          "id": "medical-01",
          "register": "medical",
          "faTokens": 31,
          "enTokens": 19,
          "ratio": 1.632
        },
        {
          "id": "instruction-01",
          "register": "instruction",
          "faTokens": 29,
          "enTokens": 23,
          "ratio": 1.261
        },
        {
          "id": "instruction-02",
          "register": "instruction",
          "faTokens": 29,
          "enTokens": 26,
          "ratio": 1.115
        }
      ],
      "faTokens": 908,
      "enTokens": 692,
      "ratio": 1.312,
      "faTokensPerChar": 0.2962,
      "enTokensPerChar": 0.1927,
      "faCharsPerToken": 3.376,
      "rank": 2,
      "tied": true
    },
    {
      "id": "mistral-small-3",
      "label": "Mistral Small 3",
      "vendor": "Mistral AI",
      "repo": "mistralai/Mistral-Small-24B-Instruct-2501",
      "mirror": false,
      "persianSpecific": false,
      "note": "Tekken — the larger vocabulary Mistral moved to.",
      "revision": "9527884be6e5616bdd54de542f9ae13384489724",
      "fileSha256": "b76085f9923309d873994d444989f7eb6ec074b06f25b58f1e8d7b7741070949",
      "measuredAt": "2026-08-19",
      "vocabSize": 131072,
      "fingerprint": "c0eabb636dbd2b86",
      "items": [
        {
          "id": "news-01",
          "register": "news",
          "faTokens": 41,
          "enTokens": 34,
          "ratio": 1.206
        },
        {
          "id": "news-02",
          "register": "news",
          "faTokens": 38,
          "enTokens": 35,
          "ratio": 1.086
        },
        {
          "id": "news-03",
          "register": "news",
          "faTokens": 36,
          "enTokens": 25,
          "ratio": 1.44
        },
        {
          "id": "technical-01",
          "register": "technical",
          "faTokens": 40,
          "enTokens": 29,
          "ratio": 1.379
        },
        {
          "id": "technical-02",
          "register": "technical",
          "faTokens": 33,
          "enTokens": 23,
          "ratio": 1.435
        },
        {
          "id": "technical-03",
          "register": "technical",
          "faTokens": 35,
          "enTokens": 24,
          "ratio": 1.458
        },
        {
          "id": "technical-04",
          "register": "technical",
          "faTokens": 38,
          "enTokens": 25,
          "ratio": 1.52
        },
        {
          "id": "policy-01",
          "register": "policy",
          "faTokens": 39,
          "enTokens": 26,
          "ratio": 1.5
        },
        {
          "id": "policy-02",
          "register": "policy",
          "faTokens": 33,
          "enTokens": 18,
          "ratio": 1.833
        },
        {
          "id": "conversational-01",
          "register": "conversational",
          "faTokens": 33,
          "enTokens": 25,
          "ratio": 1.32
        },
        {
          "id": "conversational-02",
          "register": "conversational",
          "faTokens": 29,
          "enTokens": 22,
          "ratio": 1.318
        },
        {
          "id": "conversational-03",
          "register": "conversational",
          "faTokens": 28,
          "enTokens": 25,
          "ratio": 1.12
        },
        {
          "id": "scientific-01",
          "register": "scientific",
          "faTokens": 47,
          "enTokens": 30,
          "ratio": 1.567
        },
        {
          "id": "scientific-02",
          "register": "scientific",
          "faTokens": 34,
          "enTokens": 25,
          "ratio": 1.36
        },
        {
          "id": "scientific-03",
          "register": "scientific",
          "faTokens": 36,
          "enTokens": 25,
          "ratio": 1.44
        },
        {
          "id": "product-01",
          "register": "product",
          "faTokens": 29,
          "enTokens": 24,
          "ratio": 1.208
        },
        {
          "id": "product-02",
          "register": "product",
          "faTokens": 31,
          "enTokens": 21,
          "ratio": 1.476
        },
        {
          "id": "finance-01",
          "register": "finance",
          "faTokens": 34,
          "enTokens": 24,
          "ratio": 1.417
        },
        {
          "id": "finance-02",
          "register": "finance",
          "faTokens": 32,
          "enTokens": 21,
          "ratio": 1.524
        },
        {
          "id": "education-01",
          "register": "education",
          "faTokens": 37,
          "enTokens": 27,
          "ratio": 1.37
        },
        {
          "id": "education-02",
          "register": "education",
          "faTokens": 26,
          "enTokens": 24,
          "ratio": 1.083
        },
        {
          "id": "legal-01",
          "register": "legal",
          "faTokens": 35,
          "enTokens": 31,
          "ratio": 1.129
        },
        {
          "id": "legal-02",
          "register": "legal",
          "faTokens": 24,
          "enTokens": 21,
          "ratio": 1.143
        },
        {
          "id": "literary-01",
          "register": "literary",
          "faTokens": 39,
          "enTokens": 26,
          "ratio": 1.5
        },
        {
          "id": "literary-02",
          "register": "literary",
          "faTokens": 32,
          "enTokens": 20,
          "ratio": 1.6
        },
        {
          "id": "medical-01",
          "register": "medical",
          "faTokens": 32,
          "enTokens": 19,
          "ratio": 1.684
        },
        {
          "id": "instruction-01",
          "register": "instruction",
          "faTokens": 33,
          "enTokens": 23,
          "ratio": 1.435
        },
        {
          "id": "instruction-02",
          "register": "instruction",
          "faTokens": 31,
          "enTokens": 26,
          "ratio": 1.192
        }
      ],
      "faTokens": 955,
      "enTokens": 698,
      "ratio": 1.368,
      "faTokensPerChar": 0.3116,
      "enTokensPerChar": 0.1943,
      "faCharsPerToken": 3.209,
      "rank": 4,
      "tied": false
    },
    {
      "id": "llama-3",
      "label": "Llama 3",
      "vendor": "Meta",
      "repo": "Xenova/llama-3-tokenizer",
      "mirror": true,
      "persianSpecific": false,
      "note": "Weights repo is licence-gated; the tokenizer is not.",
      "revision": "72bff9ee09897a16b3b4b2b9995fecb0bfa7dbe6",
      "fileSha256": "c05a3c2174e9edd5be19dc5a0748c42a9037bec2811ce062728bfd71f8702d78",
      "measuredAt": "2026-08-19",
      "vocabSize": 128256,
      "fingerprint": "68ca3b8ce39c96b2",
      "items": [
        {
          "id": "news-01",
          "register": "news",
          "faTokens": 40,
          "enTokens": 34,
          "ratio": 1.176
        },
        {
          "id": "news-02",
          "register": "news",
          "faTokens": 37,
          "enTokens": 34,
          "ratio": 1.088
        },
        {
          "id": "news-03",
          "register": "news",
          "faTokens": 37,
          "enTokens": 24,
          "ratio": 1.542
        },
        {
          "id": "technical-01",
          "register": "technical",
          "faTokens": 39,
          "enTokens": 29,
          "ratio": 1.345
        },
        {
          "id": "technical-02",
          "register": "technical",
          "faTokens": 35,
          "enTokens": 22,
          "ratio": 1.591
        },
        {
          "id": "technical-03",
          "register": "technical",
          "faTokens": 34,
          "enTokens": 24,
          "ratio": 1.417
        },
        {
          "id": "technical-04",
          "register": "technical",
          "faTokens": 36,
          "enTokens": 25,
          "ratio": 1.44
        },
        {
          "id": "policy-01",
          "register": "policy",
          "faTokens": 39,
          "enTokens": 25,
          "ratio": 1.56
        },
        {
          "id": "policy-02",
          "register": "policy",
          "faTokens": 36,
          "enTokens": 18,
          "ratio": 2
        },
        {
          "id": "conversational-01",
          "register": "conversational",
          "faTokens": 35,
          "enTokens": 25,
          "ratio": 1.4
        },
        {
          "id": "conversational-02",
          "register": "conversational",
          "faTokens": 29,
          "enTokens": 22,
          "ratio": 1.318
        },
        {
          "id": "conversational-03",
          "register": "conversational",
          "faTokens": 29,
          "enTokens": 25,
          "ratio": 1.16
        },
        {
          "id": "scientific-01",
          "register": "scientific",
          "faTokens": 47,
          "enTokens": 28,
          "ratio": 1.679
        },
        {
          "id": "scientific-02",
          "register": "scientific",
          "faTokens": 35,
          "enTokens": 25,
          "ratio": 1.4
        },
        {
          "id": "scientific-03",
          "register": "scientific",
          "faTokens": 36,
          "enTokens": 25,
          "ratio": 1.44
        },
        {
          "id": "product-01",
          "register": "product",
          "faTokens": 33,
          "enTokens": 24,
          "ratio": 1.375
        },
        {
          "id": "product-02",
          "register": "product",
          "faTokens": 36,
          "enTokens": 20,
          "ratio": 1.8
        },
        {
          "id": "finance-01",
          "register": "finance",
          "faTokens": 32,
          "enTokens": 24,
          "ratio": 1.333
        },
        {
          "id": "finance-02",
          "register": "finance",
          "faTokens": 30,
          "enTokens": 20,
          "ratio": 1.5
        },
        {
          "id": "education-01",
          "register": "education",
          "faTokens": 38,
          "enTokens": 26,
          "ratio": 1.462
        },
        {
          "id": "education-02",
          "register": "education",
          "faTokens": 23,
          "enTokens": 24,
          "ratio": 0.958
        },
        {
          "id": "legal-01",
          "register": "legal",
          "faTokens": 35,
          "enTokens": 31,
          "ratio": 1.129
        },
        {
          "id": "legal-02",
          "register": "legal",
          "faTokens": 27,
          "enTokens": 21,
          "ratio": 1.286
        },
        {
          "id": "literary-01",
          "register": "literary",
          "faTokens": 37,
          "enTokens": 26,
          "ratio": 1.423
        },
        {
          "id": "literary-02",
          "register": "literary",
          "faTokens": 29,
          "enTokens": 20,
          "ratio": 1.45
        },
        {
          "id": "medical-01",
          "register": "medical",
          "faTokens": 33,
          "enTokens": 19,
          "ratio": 1.737
        },
        {
          "id": "instruction-01",
          "register": "instruction",
          "faTokens": 32,
          "enTokens": 23,
          "ratio": 1.391
        },
        {
          "id": "instruction-02",
          "register": "instruction",
          "faTokens": 30,
          "enTokens": 26,
          "ratio": 1.154
        }
      ],
      "faTokens": 959,
      "enTokens": 689,
      "ratio": 1.392,
      "faTokensPerChar": 0.3129,
      "enTokensPerChar": 0.1918,
      "faCharsPerToken": 3.196,
      "rank": 5,
      "tied": true
    },
    {
      "id": "llama-3-1",
      "label": "Llama 3.1",
      "vendor": "Meta",
      "repo": "Xenova/Meta-Llama-3.1-Tokenizer",
      "mirror": true,
      "persianSpecific": false,
      "note": "Listed separately so an identical fingerprint to Llama 3 is a published finding rather than an assumption.",
      "revision": "d49ea33e96b3fcf8738bb79ff9d85da81742fa82",
      "fileSha256": "79e3e522635f3171300913bb421464a87de6222182a0570b9b2ccba2a964b2b4",
      "measuredAt": "2026-08-19",
      "vocabSize": 128256,
      "fingerprint": "68ca3b8ce39c96b2",
      "items": [
        {
          "id": "news-01",
          "register": "news",
          "faTokens": 40,
          "enTokens": 34,
          "ratio": 1.176
        },
        {
          "id": "news-02",
          "register": "news",
          "faTokens": 37,
          "enTokens": 34,
          "ratio": 1.088
        },
        {
          "id": "news-03",
          "register": "news",
          "faTokens": 37,
          "enTokens": 24,
          "ratio": 1.542
        },
        {
          "id": "technical-01",
          "register": "technical",
          "faTokens": 39,
          "enTokens": 29,
          "ratio": 1.345
        },
        {
          "id": "technical-02",
          "register": "technical",
          "faTokens": 35,
          "enTokens": 22,
          "ratio": 1.591
        },
        {
          "id": "technical-03",
          "register": "technical",
          "faTokens": 34,
          "enTokens": 24,
          "ratio": 1.417
        },
        {
          "id": "technical-04",
          "register": "technical",
          "faTokens": 36,
          "enTokens": 25,
          "ratio": 1.44
        },
        {
          "id": "policy-01",
          "register": "policy",
          "faTokens": 39,
          "enTokens": 25,
          "ratio": 1.56
        },
        {
          "id": "policy-02",
          "register": "policy",
          "faTokens": 36,
          "enTokens": 18,
          "ratio": 2
        },
        {
          "id": "conversational-01",
          "register": "conversational",
          "faTokens": 35,
          "enTokens": 25,
          "ratio": 1.4
        },
        {
          "id": "conversational-02",
          "register": "conversational",
          "faTokens": 29,
          "enTokens": 22,
          "ratio": 1.318
        },
        {
          "id": "conversational-03",
          "register": "conversational",
          "faTokens": 29,
          "enTokens": 25,
          "ratio": 1.16
        },
        {
          "id": "scientific-01",
          "register": "scientific",
          "faTokens": 47,
          "enTokens": 28,
          "ratio": 1.679
        },
        {
          "id": "scientific-02",
          "register": "scientific",
          "faTokens": 35,
          "enTokens": 25,
          "ratio": 1.4
        },
        {
          "id": "scientific-03",
          "register": "scientific",
          "faTokens": 36,
          "enTokens": 25,
          "ratio": 1.44
        },
        {
          "id": "product-01",
          "register": "product",
          "faTokens": 33,
          "enTokens": 24,
          "ratio": 1.375
        },
        {
          "id": "product-02",
          "register": "product",
          "faTokens": 36,
          "enTokens": 20,
          "ratio": 1.8
        },
        {
          "id": "finance-01",
          "register": "finance",
          "faTokens": 32,
          "enTokens": 24,
          "ratio": 1.333
        },
        {
          "id": "finance-02",
          "register": "finance",
          "faTokens": 30,
          "enTokens": 20,
          "ratio": 1.5
        },
        {
          "id": "education-01",
          "register": "education",
          "faTokens": 38,
          "enTokens": 26,
          "ratio": 1.462
        },
        {
          "id": "education-02",
          "register": "education",
          "faTokens": 23,
          "enTokens": 24,
          "ratio": 0.958
        },
        {
          "id": "legal-01",
          "register": "legal",
          "faTokens": 35,
          "enTokens": 31,
          "ratio": 1.129
        },
        {
          "id": "legal-02",
          "register": "legal",
          "faTokens": 27,
          "enTokens": 21,
          "ratio": 1.286
        },
        {
          "id": "literary-01",
          "register": "literary",
          "faTokens": 37,
          "enTokens": 26,
          "ratio": 1.423
        },
        {
          "id": "literary-02",
          "register": "literary",
          "faTokens": 29,
          "enTokens": 20,
          "ratio": 1.45
        },
        {
          "id": "medical-01",
          "register": "medical",
          "faTokens": 33,
          "enTokens": 19,
          "ratio": 1.737
        },
        {
          "id": "instruction-01",
          "register": "instruction",
          "faTokens": 32,
          "enTokens": 23,
          "ratio": 1.391
        },
        {
          "id": "instruction-02",
          "register": "instruction",
          "faTokens": 30,
          "enTokens": 26,
          "ratio": 1.154
        }
      ],
      "faTokens": 959,
      "enTokens": 689,
      "ratio": 1.392,
      "faTokensPerChar": 0.3129,
      "enTokensPerChar": 0.1918,
      "faCharsPerToken": 3.196,
      "sameTokenizerAs": "llama-3",
      "rank": 5,
      "tied": true
    },
    {
      "id": "gpt-4o",
      "label": "GPT-4o",
      "vendor": "OpenAI",
      "repo": "Xenova/gpt-4o",
      "mirror": true,
      "persianSpecific": false,
      "note": "o200k_base — also used by GPT-4.1 and o-series models.",
      "revision": "7956d98f2a83b2751a98ea7136fdf7fe6cf54e69",
      "fileSha256": "43a3ad4618a6a938f8c2614154ea10c31ad53a62d5683ae0b0e6133575cef07e",
      "measuredAt": "2026-08-19",
      "vocabSize": 200000,
      "fingerprint": "e4a1818eed0c8d85",
      "items": [
        {
          "id": "news-01",
          "register": "news",
          "faTokens": 39,
          "enTokens": 34,
          "ratio": 1.147
        },
        {
          "id": "news-02",
          "register": "news",
          "faTokens": 38,
          "enTokens": 34,
          "ratio": 1.118
        },
        {
          "id": "news-03",
          "register": "news",
          "faTokens": 39,
          "enTokens": 24,
          "ratio": 1.625
        },
        {
          "id": "technical-01",
          "register": "technical",
          "faTokens": 38,
          "enTokens": 29,
          "ratio": 1.31
        },
        {
          "id": "technical-02",
          "register": "technical",
          "faTokens": 37,
          "enTokens": 22,
          "ratio": 1.682
        },
        {
          "id": "technical-03",
          "register": "technical",
          "faTokens": 36,
          "enTokens": 24,
          "ratio": 1.5
        },
        {
          "id": "technical-04",
          "register": "technical",
          "faTokens": 36,
          "enTokens": 25,
          "ratio": 1.44
        },
        {
          "id": "policy-01",
          "register": "policy",
          "faTokens": 39,
          "enTokens": 25,
          "ratio": 1.56
        },
        {
          "id": "policy-02",
          "register": "policy",
          "faTokens": 36,
          "enTokens": 18,
          "ratio": 2
        },
        {
          "id": "conversational-01",
          "register": "conversational",
          "faTokens": 31,
          "enTokens": 25,
          "ratio": 1.24
        },
        {
          "id": "conversational-02",
          "register": "conversational",
          "faTokens": 28,
          "enTokens": 23,
          "ratio": 1.217
        },
        {
          "id": "conversational-03",
          "register": "conversational",
          "faTokens": 29,
          "enTokens": 25,
          "ratio": 1.16
        },
        {
          "id": "scientific-01",
          "register": "scientific",
          "faTokens": 47,
          "enTokens": 28,
          "ratio": 1.679
        },
        {
          "id": "scientific-02",
          "register": "scientific",
          "faTokens": 35,
          "enTokens": 25,
          "ratio": 1.4
        },
        {
          "id": "scientific-03",
          "register": "scientific",
          "faTokens": 37,
          "enTokens": 25,
          "ratio": 1.48
        },
        {
          "id": "product-01",
          "register": "product",
          "faTokens": 33,
          "enTokens": 24,
          "ratio": 1.375
        },
        {
          "id": "product-02",
          "register": "product",
          "faTokens": 35,
          "enTokens": 20,
          "ratio": 1.75
        },
        {
          "id": "finance-01",
          "register": "finance",
          "faTokens": 32,
          "enTokens": 24,
          "ratio": 1.333
        },
        {
          "id": "finance-02",
          "register": "finance",
          "faTokens": 31,
          "enTokens": 20,
          "ratio": 1.55
        },
        {
          "id": "education-01",
          "register": "education",
          "faTokens": 36,
          "enTokens": 26,
          "ratio": 1.385
        },
        {
          "id": "education-02",
          "register": "education",
          "faTokens": 26,
          "enTokens": 24,
          "ratio": 1.083
        },
        {
          "id": "legal-01",
          "register": "legal",
          "faTokens": 34,
          "enTokens": 30,
          "ratio": 1.133
        },
        {
          "id": "legal-02",
          "register": "legal",
          "faTokens": 25,
          "enTokens": 21,
          "ratio": 1.19
        },
        {
          "id": "literary-01",
          "register": "literary",
          "faTokens": 38,
          "enTokens": 26,
          "ratio": 1.462
        },
        {
          "id": "literary-02",
          "register": "literary",
          "faTokens": 34,
          "enTokens": 20,
          "ratio": 1.7
        },
        {
          "id": "medical-01",
          "register": "medical",
          "faTokens": 33,
          "enTokens": 19,
          "ratio": 1.737
        },
        {
          "id": "instruction-01",
          "register": "instruction",
          "faTokens": 34,
          "enTokens": 23,
          "ratio": 1.478
        },
        {
          "id": "instruction-02",
          "register": "instruction",
          "faTokens": 33,
          "enTokens": 26,
          "ratio": 1.269
        }
      ],
      "faTokens": 969,
      "enTokens": 689,
      "ratio": 1.406,
      "faTokensPerChar": 0.3162,
      "enTokensPerChar": 0.1918,
      "faCharsPerToken": 3.163,
      "rank": 7,
      "tied": true
    },
    {
      "id": "gpt-oss-20b",
      "label": "gpt-oss 20B",
      "vendor": "OpenAI",
      "repo": "openai/gpt-oss-20b",
      "mirror": false,
      "persianSpecific": false,
      "note": "Open-weights, published by the vendor.",
      "revision": "6cee5e81ee83917806bbde320786a8fb61efebee",
      "fileSha256": "0614fe83cadab421296e664e1f48f4261fa8fef6e03e63bb75c20f38e37d07d3",
      "measuredAt": "2026-08-19",
      "vocabSize": 200019,
      "fingerprint": "f5c4cab9908b0d33",
      "items": [
        {
          "id": "news-01",
          "register": "news",
          "faTokens": 39,
          "enTokens": 34,
          "ratio": 1.147
        },
        {
          "id": "news-02",
          "register": "news",
          "faTokens": 38,
          "enTokens": 34,
          "ratio": 1.118
        },
        {
          "id": "news-03",
          "register": "news",
          "faTokens": 39,
          "enTokens": 24,
          "ratio": 1.625
        },
        {
          "id": "technical-01",
          "register": "technical",
          "faTokens": 38,
          "enTokens": 29,
          "ratio": 1.31
        },
        {
          "id": "technical-02",
          "register": "technical",
          "faTokens": 37,
          "enTokens": 22,
          "ratio": 1.682
        },
        {
          "id": "technical-03",
          "register": "technical",
          "faTokens": 36,
          "enTokens": 24,
          "ratio": 1.5
        },
        {
          "id": "technical-04",
          "register": "technical",
          "faTokens": 36,
          "enTokens": 25,
          "ratio": 1.44
        },
        {
          "id": "policy-01",
          "register": "policy",
          "faTokens": 39,
          "enTokens": 25,
          "ratio": 1.56
        },
        {
          "id": "policy-02",
          "register": "policy",
          "faTokens": 36,
          "enTokens": 18,
          "ratio": 2
        },
        {
          "id": "conversational-01",
          "register": "conversational",
          "faTokens": 31,
          "enTokens": 25,
          "ratio": 1.24
        },
        {
          "id": "conversational-02",
          "register": "conversational",
          "faTokens": 28,
          "enTokens": 23,
          "ratio": 1.217
        },
        {
          "id": "conversational-03",
          "register": "conversational",
          "faTokens": 29,
          "enTokens": 25,
          "ratio": 1.16
        },
        {
          "id": "scientific-01",
          "register": "scientific",
          "faTokens": 47,
          "enTokens": 28,
          "ratio": 1.679
        },
        {
          "id": "scientific-02",
          "register": "scientific",
          "faTokens": 35,
          "enTokens": 25,
          "ratio": 1.4
        },
        {
          "id": "scientific-03",
          "register": "scientific",
          "faTokens": 37,
          "enTokens": 25,
          "ratio": 1.48
        },
        {
          "id": "product-01",
          "register": "product",
          "faTokens": 33,
          "enTokens": 24,
          "ratio": 1.375
        },
        {
          "id": "product-02",
          "register": "product",
          "faTokens": 35,
          "enTokens": 20,
          "ratio": 1.75
        },
        {
          "id": "finance-01",
          "register": "finance",
          "faTokens": 32,
          "enTokens": 24,
          "ratio": 1.333
        },
        {
          "id": "finance-02",
          "register": "finance",
          "faTokens": 31,
          "enTokens": 20,
          "ratio": 1.55
        },
        {
          "id": "education-01",
          "register": "education",
          "faTokens": 36,
          "enTokens": 26,
          "ratio": 1.385
        },
        {
          "id": "education-02",
          "register": "education",
          "faTokens": 26,
          "enTokens": 24,
          "ratio": 1.083
        },
        {
          "id": "legal-01",
          "register": "legal",
          "faTokens": 34,
          "enTokens": 30,
          "ratio": 1.133
        },
        {
          "id": "legal-02",
          "register": "legal",
          "faTokens": 25,
          "enTokens": 21,
          "ratio": 1.19
        },
        {
          "id": "literary-01",
          "register": "literary",
          "faTokens": 38,
          "enTokens": 26,
          "ratio": 1.462
        },
        {
          "id": "literary-02",
          "register": "literary",
          "faTokens": 34,
          "enTokens": 20,
          "ratio": 1.7
        },
        {
          "id": "medical-01",
          "register": "medical",
          "faTokens": 33,
          "enTokens": 19,
          "ratio": 1.737
        },
        {
          "id": "instruction-01",
          "register": "instruction",
          "faTokens": 34,
          "enTokens": 23,
          "ratio": 1.478
        },
        {
          "id": "instruction-02",
          "register": "instruction",
          "faTokens": 33,
          "enTokens": 26,
          "ratio": 1.269
        }
      ],
      "faTokens": 969,
      "enTokens": 689,
      "ratio": 1.406,
      "faTokensPerChar": 0.3162,
      "enTokensPerChar": 0.1918,
      "faCharsPerToken": 3.163,
      "rank": 7,
      "tied": true
    },
    {
      "id": "glm-4-5",
      "label": "GLM-4.5",
      "vendor": "Z.ai",
      "repo": "zai-org/GLM-4.5",
      "mirror": false,
      "persianSpecific": false,
      "note": "",
      "revision": "cbb2c7cfb52fa128a9660cb1a7a78e017899e115",
      "fileSha256": "9340665016419c825c4bdabbcc9acc43b7ca2c68ce142724afa829abb1be5efd",
      "measuredAt": "2026-08-19",
      "vocabSize": 151365,
      "fingerprint": "3c4bed0010b8e7ab",
      "items": [
        {
          "id": "news-01",
          "register": "news",
          "faTokens": 51,
          "enTokens": 34,
          "ratio": 1.5
        },
        {
          "id": "news-02",
          "register": "news",
          "faTokens": 47,
          "enTokens": 34,
          "ratio": 1.382
        },
        {
          "id": "news-03",
          "register": "news",
          "faTokens": 46,
          "enTokens": 24,
          "ratio": 1.917
        },
        {
          "id": "technical-01",
          "register": "technical",
          "faTokens": 51,
          "enTokens": 29,
          "ratio": 1.759
        },
        {
          "id": "technical-02",
          "register": "technical",
          "faTokens": 44,
          "enTokens": 22,
          "ratio": 2
        },
        {
          "id": "technical-03",
          "register": "technical",
          "faTokens": 46,
          "enTokens": 24,
          "ratio": 1.917
        },
        {
          "id": "technical-04",
          "register": "technical",
          "faTokens": 46,
          "enTokens": 25,
          "ratio": 1.84
        },
        {
          "id": "policy-01",
          "register": "policy",
          "faTokens": 47,
          "enTokens": 25,
          "ratio": 1.88
        },
        {
          "id": "policy-02",
          "register": "policy",
          "faTokens": 45,
          "enTokens": 18,
          "ratio": 2.5
        },
        {
          "id": "conversational-01",
          "register": "conversational",
          "faTokens": 45,
          "enTokens": 25,
          "ratio": 1.8
        },
        {
          "id": "conversational-02",
          "register": "conversational",
          "faTokens": 35,
          "enTokens": 22,
          "ratio": 1.591
        },
        {
          "id": "conversational-03",
          "register": "conversational",
          "faTokens": 35,
          "enTokens": 25,
          "ratio": 1.4
        },
        {
          "id": "scientific-01",
          "register": "scientific",
          "faTokens": 53,
          "enTokens": 28,
          "ratio": 1.893
        },
        {
          "id": "scientific-02",
          "register": "scientific",
          "faTokens": 43,
          "enTokens": 25,
          "ratio": 1.72
        },
        {
          "id": "scientific-03",
          "register": "scientific",
          "faTokens": 46,
          "enTokens": 25,
          "ratio": 1.84
        },
        {
          "id": "product-01",
          "register": "product",
          "faTokens": 35,
          "enTokens": 24,
          "ratio": 1.458
        },
        {
          "id": "product-02",
          "register": "product",
          "faTokens": 41,
          "enTokens": 20,
          "ratio": 2.05
        },
        {
          "id": "finance-01",
          "register": "finance",
          "faTokens": 42,
          "enTokens": 24,
          "ratio": 1.75
        },
        {
          "id": "finance-02",
          "register": "finance",
          "faTokens": 35,
          "enTokens": 20,
          "ratio": 1.75
        },
        {
          "id": "education-01",
          "register": "education",
          "faTokens": 49,
          "enTokens": 26,
          "ratio": 1.885
        },
        {
          "id": "education-02",
          "register": "education",
          "faTokens": 32,
          "enTokens": 24,
          "ratio": 1.333
        },
        {
          "id": "legal-01",
          "register": "legal",
          "faTokens": 45,
          "enTokens": 31,
          "ratio": 1.452
        },
        {
          "id": "legal-02",
          "register": "legal",
          "faTokens": 29,
          "enTokens": 21,
          "ratio": 1.381
        },
        {
          "id": "literary-01",
          "register": "literary",
          "faTokens": 46,
          "enTokens": 26,
          "ratio": 1.769
        },
        {
          "id": "literary-02",
          "register": "literary",
          "faTokens": 39,
          "enTokens": 20,
          "ratio": 1.95
        },
        {
          "id": "medical-01",
          "register": "medical",
          "faTokens": 37,
          "enTokens": 19,
          "ratio": 1.947
        },
        {
          "id": "instruction-01",
          "register": "instruction",
          "faTokens": 45,
          "enTokens": 23,
          "ratio": 1.957
        },
        {
          "id": "instruction-02",
          "register": "instruction",
          "faTokens": 36,
          "enTokens": 26,
          "ratio": 1.385
        }
      ],
      "faTokens": 1191,
      "enTokens": 689,
      "ratio": 1.729,
      "faTokensPerChar": 0.3886,
      "enTokensPerChar": 0.1918,
      "faCharsPerToken": 2.573,
      "rank": 9,
      "tied": false
    },
    {
      "id": "deepseek-v3",
      "label": "DeepSeek-V3",
      "vendor": "DeepSeek",
      "repo": "deepseek-ai/DeepSeek-V3",
      "mirror": false,
      "persianSpecific": false,
      "note": "",
      "revision": "e815299b0bcbac849fa540c768ef21845365c9eb",
      "fileSha256": "621ac2e32d0dba658404412318818aaa8ce8cda492e59830109d8da6b517fb41",
      "measuredAt": "2026-08-19",
      "vocabSize": 128815,
      "fingerprint": "b49c0fe0d563eb09",
      "items": [
        {
          "id": "news-01",
          "register": "news",
          "faTokens": 54,
          "enTokens": 34,
          "ratio": 1.588
        },
        {
          "id": "news-02",
          "register": "news",
          "faTokens": 46,
          "enTokens": 34,
          "ratio": 1.353
        },
        {
          "id": "news-03",
          "register": "news",
          "faTokens": 49,
          "enTokens": 24,
          "ratio": 2.042
        },
        {
          "id": "technical-01",
          "register": "technical",
          "faTokens": 51,
          "enTokens": 29,
          "ratio": 1.759
        },
        {
          "id": "technical-02",
          "register": "technical",
          "faTokens": 42,
          "enTokens": 23,
          "ratio": 1.826
        },
        {
          "id": "technical-03",
          "register": "technical",
          "faTokens": 50,
          "enTokens": 24,
          "ratio": 2.083
        },
        {
          "id": "technical-04",
          "register": "technical",
          "faTokens": 49,
          "enTokens": 25,
          "ratio": 1.96
        },
        {
          "id": "policy-01",
          "register": "policy",
          "faTokens": 55,
          "enTokens": 26,
          "ratio": 2.115
        },
        {
          "id": "policy-02",
          "register": "policy",
          "faTokens": 48,
          "enTokens": 18,
          "ratio": 2.667
        },
        {
          "id": "conversational-01",
          "register": "conversational",
          "faTokens": 42,
          "enTokens": 25,
          "ratio": 1.68
        },
        {
          "id": "conversational-02",
          "register": "conversational",
          "faTokens": 37,
          "enTokens": 22,
          "ratio": 1.682
        },
        {
          "id": "conversational-03",
          "register": "conversational",
          "faTokens": 35,
          "enTokens": 25,
          "ratio": 1.4
        },
        {
          "id": "scientific-01",
          "register": "scientific",
          "faTokens": 56,
          "enTokens": 30,
          "ratio": 1.867
        },
        {
          "id": "scientific-02",
          "register": "scientific",
          "faTokens": 40,
          "enTokens": 25,
          "ratio": 1.6
        },
        {
          "id": "scientific-03",
          "register": "scientific",
          "faTokens": 52,
          "enTokens": 25,
          "ratio": 2.08
        },
        {
          "id": "product-01",
          "register": "product",
          "faTokens": 34,
          "enTokens": 24,
          "ratio": 1.417
        },
        {
          "id": "product-02",
          "register": "product",
          "faTokens": 43,
          "enTokens": 20,
          "ratio": 2.15
        },
        {
          "id": "finance-01",
          "register": "finance",
          "faTokens": 42,
          "enTokens": 24,
          "ratio": 1.75
        },
        {
          "id": "finance-02",
          "register": "finance",
          "faTokens": 33,
          "enTokens": 20,
          "ratio": 1.65
        },
        {
          "id": "education-01",
          "register": "education",
          "faTokens": 44,
          "enTokens": 26,
          "ratio": 1.692
        },
        {
          "id": "education-02",
          "register": "education",
          "faTokens": 29,
          "enTokens": 24,
          "ratio": 1.208
        },
        {
          "id": "legal-01",
          "register": "legal",
          "faTokens": 43,
          "enTokens": 31,
          "ratio": 1.387
        },
        {
          "id": "legal-02",
          "register": "legal",
          "faTokens": 29,
          "enTokens": 21,
          "ratio": 1.381
        },
        {
          "id": "literary-01",
          "register": "literary",
          "faTokens": 48,
          "enTokens": 26,
          "ratio": 1.846
        },
        {
          "id": "literary-02",
          "register": "literary",
          "faTokens": 40,
          "enTokens": 20,
          "ratio": 2
        },
        {
          "id": "medical-01",
          "register": "medical",
          "faTokens": 35,
          "enTokens": 19,
          "ratio": 1.842
        },
        {
          "id": "instruction-01",
          "register": "instruction",
          "faTokens": 42,
          "enTokens": 23,
          "ratio": 1.826
        },
        {
          "id": "instruction-02",
          "register": "instruction",
          "faTokens": 35,
          "enTokens": 26,
          "ratio": 1.346
        }
      ],
      "faTokens": 1203,
      "enTokens": 693,
      "ratio": 1.736,
      "faTokensPerChar": 0.3925,
      "enTokensPerChar": 0.1929,
      "faCharsPerToken": 2.548,
      "rank": 10,
      "tied": false
    },
    {
      "id": "qwen-2-5",
      "label": "Qwen2.5",
      "vendor": "Alibaba",
      "repo": "Qwen/Qwen2.5-7B-Instruct",
      "mirror": false,
      "persianSpecific": false,
      "note": "",
      "revision": "a09a35458c702b33eeacc393d103063234e8bc28",
      "fileSha256": "c0382117ea329cdf097041132f6d735924b697924d6f6fc3945713e96ce87539",
      "measuredAt": "2026-08-19",
      "vocabSize": 151665,
      "fingerprint": "ab095f79e58ac767",
      "items": [
        {
          "id": "news-01",
          "register": "news",
          "faTokens": 87,
          "enTokens": 34,
          "ratio": 2.559
        },
        {
          "id": "news-02",
          "register": "news",
          "faTokens": 80,
          "enTokens": 34,
          "ratio": 2.353
        },
        {
          "id": "news-03",
          "register": "news",
          "faTokens": 79,
          "enTokens": 24,
          "ratio": 3.292
        },
        {
          "id": "technical-01",
          "register": "technical",
          "faTokens": 81,
          "enTokens": 29,
          "ratio": 2.793
        },
        {
          "id": "technical-02",
          "register": "technical",
          "faTokens": 68,
          "enTokens": 22,
          "ratio": 3.091
        },
        {
          "id": "technical-03",
          "register": "technical",
          "faTokens": 69,
          "enTokens": 24,
          "ratio": 2.875
        },
        {
          "id": "technical-04",
          "register": "technical",
          "faTokens": 67,
          "enTokens": 25,
          "ratio": 2.68
        },
        {
          "id": "policy-01",
          "register": "policy",
          "faTokens": 69,
          "enTokens": 25,
          "ratio": 2.76
        },
        {
          "id": "policy-02",
          "register": "policy",
          "faTokens": 65,
          "enTokens": 18,
          "ratio": 3.611
        },
        {
          "id": "conversational-01",
          "register": "conversational",
          "faTokens": 55,
          "enTokens": 25,
          "ratio": 2.2
        },
        {
          "id": "conversational-02",
          "register": "conversational",
          "faTokens": 46,
          "enTokens": 22,
          "ratio": 2.091
        },
        {
          "id": "conversational-03",
          "register": "conversational",
          "faTokens": 53,
          "enTokens": 25,
          "ratio": 2.12
        },
        {
          "id": "scientific-01",
          "register": "scientific",
          "faTokens": 81,
          "enTokens": 28,
          "ratio": 2.893
        },
        {
          "id": "scientific-02",
          "register": "scientific",
          "faTokens": 56,
          "enTokens": 25,
          "ratio": 2.24
        },
        {
          "id": "scientific-03",
          "register": "scientific",
          "faTokens": 75,
          "enTokens": 25,
          "ratio": 3
        },
        {
          "id": "product-01",
          "register": "product",
          "faTokens": 59,
          "enTokens": 24,
          "ratio": 2.458
        },
        {
          "id": "product-02",
          "register": "product",
          "faTokens": 63,
          "enTokens": 20,
          "ratio": 3.15
        },
        {
          "id": "finance-01",
          "register": "finance",
          "faTokens": 62,
          "enTokens": 24,
          "ratio": 2.583
        },
        {
          "id": "finance-02",
          "register": "finance",
          "faTokens": 60,
          "enTokens": 20,
          "ratio": 3
        },
        {
          "id": "education-01",
          "register": "education",
          "faTokens": 66,
          "enTokens": 26,
          "ratio": 2.538
        },
        {
          "id": "education-02",
          "register": "education",
          "faTokens": 50,
          "enTokens": 24,
          "ratio": 2.083
        },
        {
          "id": "legal-01",
          "register": "legal",
          "faTokens": 66,
          "enTokens": 31,
          "ratio": 2.129
        },
        {
          "id": "legal-02",
          "register": "legal",
          "faTokens": 40,
          "enTokens": 21,
          "ratio": 1.905
        },
        {
          "id": "literary-01",
          "register": "literary",
          "faTokens": 63,
          "enTokens": 26,
          "ratio": 2.423
        },
        {
          "id": "literary-02",
          "register": "literary",
          "faTokens": 49,
          "enTokens": 20,
          "ratio": 2.45
        },
        {
          "id": "medical-01",
          "register": "medical",
          "faTokens": 56,
          "enTokens": 19,
          "ratio": 2.947
        },
        {
          "id": "instruction-01",
          "register": "instruction",
          "faTokens": 63,
          "enTokens": 23,
          "ratio": 2.739
        },
        {
          "id": "instruction-02",
          "register": "instruction",
          "faTokens": 49,
          "enTokens": 26,
          "ratio": 1.885
        }
      ],
      "faTokens": 1777,
      "enTokens": 689,
      "ratio": 2.579,
      "faTokensPerChar": 0.5798,
      "enTokensPerChar": 0.1918,
      "faCharsPerToken": 1.725,
      "rank": 11,
      "tied": true
    },
    {
      "id": "qwen-3",
      "label": "Qwen3",
      "vendor": "Alibaba",
      "repo": "Qwen/Qwen3-8B",
      "mirror": false,
      "persianSpecific": false,
      "note": "",
      "revision": "b968826d9c46dd6066d109eabc6255188de91218",
      "fileSha256": "aeb13307a71acd8fe81861d94ad54ab689df773318809eed3cbe794b4492dae4",
      "measuredAt": "2026-08-19",
      "vocabSize": 151669,
      "fingerprint": "ab095f79e58ac767",
      "items": [
        {
          "id": "news-01",
          "register": "news",
          "faTokens": 87,
          "enTokens": 34,
          "ratio": 2.559
        },
        {
          "id": "news-02",
          "register": "news",
          "faTokens": 80,
          "enTokens": 34,
          "ratio": 2.353
        },
        {
          "id": "news-03",
          "register": "news",
          "faTokens": 79,
          "enTokens": 24,
          "ratio": 3.292
        },
        {
          "id": "technical-01",
          "register": "technical",
          "faTokens": 81,
          "enTokens": 29,
          "ratio": 2.793
        },
        {
          "id": "technical-02",
          "register": "technical",
          "faTokens": 68,
          "enTokens": 22,
          "ratio": 3.091
        },
        {
          "id": "technical-03",
          "register": "technical",
          "faTokens": 69,
          "enTokens": 24,
          "ratio": 2.875
        },
        {
          "id": "technical-04",
          "register": "technical",
          "faTokens": 67,
          "enTokens": 25,
          "ratio": 2.68
        },
        {
          "id": "policy-01",
          "register": "policy",
          "faTokens": 69,
          "enTokens": 25,
          "ratio": 2.76
        },
        {
          "id": "policy-02",
          "register": "policy",
          "faTokens": 65,
          "enTokens": 18,
          "ratio": 3.611
        },
        {
          "id": "conversational-01",
          "register": "conversational",
          "faTokens": 55,
          "enTokens": 25,
          "ratio": 2.2
        },
        {
          "id": "conversational-02",
          "register": "conversational",
          "faTokens": 46,
          "enTokens": 22,
          "ratio": 2.091
        },
        {
          "id": "conversational-03",
          "register": "conversational",
          "faTokens": 53,
          "enTokens": 25,
          "ratio": 2.12
        },
        {
          "id": "scientific-01",
          "register": "scientific",
          "faTokens": 81,
          "enTokens": 28,
          "ratio": 2.893
        },
        {
          "id": "scientific-02",
          "register": "scientific",
          "faTokens": 56,
          "enTokens": 25,
          "ratio": 2.24
        },
        {
          "id": "scientific-03",
          "register": "scientific",
          "faTokens": 75,
          "enTokens": 25,
          "ratio": 3
        },
        {
          "id": "product-01",
          "register": "product",
          "faTokens": 59,
          "enTokens": 24,
          "ratio": 2.458
        },
        {
          "id": "product-02",
          "register": "product",
          "faTokens": 63,
          "enTokens": 20,
          "ratio": 3.15
        },
        {
          "id": "finance-01",
          "register": "finance",
          "faTokens": 62,
          "enTokens": 24,
          "ratio": 2.583
        },
        {
          "id": "finance-02",
          "register": "finance",
          "faTokens": 60,
          "enTokens": 20,
          "ratio": 3
        },
        {
          "id": "education-01",
          "register": "education",
          "faTokens": 66,
          "enTokens": 26,
          "ratio": 2.538
        },
        {
          "id": "education-02",
          "register": "education",
          "faTokens": 50,
          "enTokens": 24,
          "ratio": 2.083
        },
        {
          "id": "legal-01",
          "register": "legal",
          "faTokens": 66,
          "enTokens": 31,
          "ratio": 2.129
        },
        {
          "id": "legal-02",
          "register": "legal",
          "faTokens": 40,
          "enTokens": 21,
          "ratio": 1.905
        },
        {
          "id": "literary-01",
          "register": "literary",
          "faTokens": 63,
          "enTokens": 26,
          "ratio": 2.423
        },
        {
          "id": "literary-02",
          "register": "literary",
          "faTokens": 49,
          "enTokens": 20,
          "ratio": 2.45
        },
        {
          "id": "medical-01",
          "register": "medical",
          "faTokens": 56,
          "enTokens": 19,
          "ratio": 2.947
        },
        {
          "id": "instruction-01",
          "register": "instruction",
          "faTokens": 63,
          "enTokens": 23,
          "ratio": 2.739
        },
        {
          "id": "instruction-02",
          "register": "instruction",
          "faTokens": 49,
          "enTokens": 26,
          "ratio": 1.885
        }
      ],
      "faTokens": 1777,
      "enTokens": 689,
      "ratio": 2.579,
      "faTokensPerChar": 0.5798,
      "enTokensPerChar": 0.1918,
      "faCharsPerToken": 1.725,
      "sameTokenizerAs": "qwen-2-5",
      "rank": 11,
      "tied": true
    },
    {
      "id": "gpt-4",
      "label": "GPT-4 / GPT-3.5 Turbo",
      "vendor": "OpenAI",
      "repo": "Xenova/gpt-4",
      "mirror": true,
      "persianSpecific": false,
      "note": "cl100k_base — the previous generation, kept for the size of the change.",
      "revision": "1d9f1f1b1fae88c0e4df1dab0a397f8de6229075",
      "fileSha256": "239eb2359f79c38497476671aaa835e01fb43d42743c612a8514a0dfa2ac93a2",
      "measuredAt": "2026-08-19",
      "vocabSize": 100263,
      "fingerprint": "390a47963496e4ba",
      "items": [
        {
          "id": "news-01",
          "register": "news",
          "faTokens": 104,
          "enTokens": 34,
          "ratio": 3.059
        },
        {
          "id": "news-02",
          "register": "news",
          "faTokens": 108,
          "enTokens": 34,
          "ratio": 3.176
        },
        {
          "id": "news-03",
          "register": "news",
          "faTokens": 96,
          "enTokens": 24,
          "ratio": 4
        },
        {
          "id": "technical-01",
          "register": "technical",
          "faTokens": 94,
          "enTokens": 29,
          "ratio": 3.241
        },
        {
          "id": "technical-02",
          "register": "technical",
          "faTokens": 77,
          "enTokens": 22,
          "ratio": 3.5
        },
        {
          "id": "technical-03",
          "register": "technical",
          "faTokens": 85,
          "enTokens": 24,
          "ratio": 3.542
        },
        {
          "id": "technical-04",
          "register": "technical",
          "faTokens": 78,
          "enTokens": 25,
          "ratio": 3.12
        },
        {
          "id": "policy-01",
          "register": "policy",
          "faTokens": 88,
          "enTokens": 25,
          "ratio": 3.52
        },
        {
          "id": "policy-02",
          "register": "policy",
          "faTokens": 82,
          "enTokens": 18,
          "ratio": 4.556
        },
        {
          "id": "conversational-01",
          "register": "conversational",
          "faTokens": 73,
          "enTokens": 25,
          "ratio": 2.92
        },
        {
          "id": "conversational-02",
          "register": "conversational",
          "faTokens": 64,
          "enTokens": 22,
          "ratio": 2.909
        },
        {
          "id": "conversational-03",
          "register": "conversational",
          "faTokens": 66,
          "enTokens": 25,
          "ratio": 2.64
        },
        {
          "id": "scientific-01",
          "register": "scientific",
          "faTokens": 100,
          "enTokens": 28,
          "ratio": 3.571
        },
        {
          "id": "scientific-02",
          "register": "scientific",
          "faTokens": 71,
          "enTokens": 25,
          "ratio": 2.84
        },
        {
          "id": "scientific-03",
          "register": "scientific",
          "faTokens": 90,
          "enTokens": 25,
          "ratio": 3.6
        },
        {
          "id": "product-01",
          "register": "product",
          "faTokens": 67,
          "enTokens": 24,
          "ratio": 2.792
        },
        {
          "id": "product-02",
          "register": "product",
          "faTokens": 72,
          "enTokens": 20,
          "ratio": 3.6
        },
        {
          "id": "finance-01",
          "register": "finance",
          "faTokens": 80,
          "enTokens": 24,
          "ratio": 3.333
        },
        {
          "id": "finance-02",
          "register": "finance",
          "faTokens": 77,
          "enTokens": 20,
          "ratio": 3.85
        },
        {
          "id": "education-01",
          "register": "education",
          "faTokens": 85,
          "enTokens": 26,
          "ratio": 3.269
        },
        {
          "id": "education-02",
          "register": "education",
          "faTokens": 61,
          "enTokens": 24,
          "ratio": 2.542
        },
        {
          "id": "legal-01",
          "register": "legal",
          "faTokens": 85,
          "enTokens": 31,
          "ratio": 2.742
        },
        {
          "id": "legal-02",
          "register": "legal",
          "faTokens": 57,
          "enTokens": 21,
          "ratio": 2.714
        },
        {
          "id": "literary-01",
          "register": "literary",
          "faTokens": 83,
          "enTokens": 26,
          "ratio": 3.192
        },
        {
          "id": "literary-02",
          "register": "literary",
          "faTokens": 66,
          "enTokens": 20,
          "ratio": 3.3
        },
        {
          "id": "medical-01",
          "register": "medical",
          "faTokens": 71,
          "enTokens": 19,
          "ratio": 3.737
        },
        {
          "id": "instruction-01",
          "register": "instruction",
          "faTokens": 79,
          "enTokens": 23,
          "ratio": 3.435
        },
        {
          "id": "instruction-02",
          "register": "instruction",
          "faTokens": 57,
          "enTokens": 26,
          "ratio": 2.192
        }
      ],
      "faTokens": 2216,
      "enTokens": 689,
      "ratio": 3.216,
      "faTokensPerChar": 0.723,
      "enTokensPerChar": 0.1918,
      "faCharsPerToken": 1.383,
      "rank": 13,
      "tied": true
    },
    {
      "id": "phi-4",
      "label": "Phi-4",
      "vendor": "Microsoft",
      "repo": "microsoft/phi-4",
      "mirror": false,
      "persianSpecific": false,
      "note": "",
      "revision": "2db69c1c3e91a05d2c64a3185acfbaf36f744e25",
      "fileSha256": "9f38d05d9d25756bb2f181ab5a0cebcd59e638df10336fc7ed1010f7296d0298",
      "measuredAt": "2026-08-19",
      "vocabSize": 100352,
      "fingerprint": "66e60352a989976f",
      "items": [
        {
          "id": "news-01",
          "register": "news",
          "faTokens": 104,
          "enTokens": 34,
          "ratio": 3.059
        },
        {
          "id": "news-02",
          "register": "news",
          "faTokens": 108,
          "enTokens": 34,
          "ratio": 3.176
        },
        {
          "id": "news-03",
          "register": "news",
          "faTokens": 96,
          "enTokens": 24,
          "ratio": 4
        },
        {
          "id": "technical-01",
          "register": "technical",
          "faTokens": 94,
          "enTokens": 29,
          "ratio": 3.241
        },
        {
          "id": "technical-02",
          "register": "technical",
          "faTokens": 77,
          "enTokens": 22,
          "ratio": 3.5
        },
        {
          "id": "technical-03",
          "register": "technical",
          "faTokens": 85,
          "enTokens": 24,
          "ratio": 3.542
        },
        {
          "id": "technical-04",
          "register": "technical",
          "faTokens": 78,
          "enTokens": 25,
          "ratio": 3.12
        },
        {
          "id": "policy-01",
          "register": "policy",
          "faTokens": 88,
          "enTokens": 25,
          "ratio": 3.52
        },
        {
          "id": "policy-02",
          "register": "policy",
          "faTokens": 82,
          "enTokens": 18,
          "ratio": 4.556
        },
        {
          "id": "conversational-01",
          "register": "conversational",
          "faTokens": 73,
          "enTokens": 25,
          "ratio": 2.92
        },
        {
          "id": "conversational-02",
          "register": "conversational",
          "faTokens": 64,
          "enTokens": 22,
          "ratio": 2.909
        },
        {
          "id": "conversational-03",
          "register": "conversational",
          "faTokens": 66,
          "enTokens": 25,
          "ratio": 2.64
        },
        {
          "id": "scientific-01",
          "register": "scientific",
          "faTokens": 100,
          "enTokens": 28,
          "ratio": 3.571
        },
        {
          "id": "scientific-02",
          "register": "scientific",
          "faTokens": 71,
          "enTokens": 25,
          "ratio": 2.84
        },
        {
          "id": "scientific-03",
          "register": "scientific",
          "faTokens": 90,
          "enTokens": 25,
          "ratio": 3.6
        },
        {
          "id": "product-01",
          "register": "product",
          "faTokens": 67,
          "enTokens": 24,
          "ratio": 2.792
        },
        {
          "id": "product-02",
          "register": "product",
          "faTokens": 72,
          "enTokens": 20,
          "ratio": 3.6
        },
        {
          "id": "finance-01",
          "register": "finance",
          "faTokens": 80,
          "enTokens": 24,
          "ratio": 3.333
        },
        {
          "id": "finance-02",
          "register": "finance",
          "faTokens": 77,
          "enTokens": 20,
          "ratio": 3.85
        },
        {
          "id": "education-01",
          "register": "education",
          "faTokens": 85,
          "enTokens": 26,
          "ratio": 3.269
        },
        {
          "id": "education-02",
          "register": "education",
          "faTokens": 61,
          "enTokens": 24,
          "ratio": 2.542
        },
        {
          "id": "legal-01",
          "register": "legal",
          "faTokens": 85,
          "enTokens": 31,
          "ratio": 2.742
        },
        {
          "id": "legal-02",
          "register": "legal",
          "faTokens": 57,
          "enTokens": 21,
          "ratio": 2.714
        },
        {
          "id": "literary-01",
          "register": "literary",
          "faTokens": 83,
          "enTokens": 26,
          "ratio": 3.192
        },
        {
          "id": "literary-02",
          "register": "literary",
          "faTokens": 66,
          "enTokens": 20,
          "ratio": 3.3
        },
        {
          "id": "medical-01",
          "register": "medical",
          "faTokens": 71,
          "enTokens": 19,
          "ratio": 3.737
        },
        {
          "id": "instruction-01",
          "register": "instruction",
          "faTokens": 79,
          "enTokens": 23,
          "ratio": 3.435
        },
        {
          "id": "instruction-02",
          "register": "instruction",
          "faTokens": 57,
          "enTokens": 26,
          "ratio": 2.192
        }
      ],
      "faTokens": 2216,
      "enTokens": 689,
      "ratio": 3.216,
      "faTokensPerChar": 0.723,
      "enTokensPerChar": 0.1918,
      "faCharsPerToken": 1.383,
      "rank": 13,
      "tied": true
    },
    {
      "id": "olmo-2",
      "label": "OLMo 2",
      "vendor": "Allen Institute for AI",
      "repo": "allenai/OLMo-2-1124-7B",
      "mirror": false,
      "persianSpecific": false,
      "note": "Fully open training data, which makes its Persian coverage inspectable.",
      "revision": "7df9a82518afdecae4e8c026b27adccc8c1f0032",
      "fileSha256": "73fd5254624f39a88e3faac6a8e11300fc3c735ed37880d4f4f08db898eaecca",
      "measuredAt": "2026-08-19",
      "vocabSize": 100278,
      "fingerprint": "ab742b330b7b43f7",
      "items": [
        {
          "id": "news-01",
          "register": "news",
          "faTokens": 104,
          "enTokens": 34,
          "ratio": 3.059
        },
        {
          "id": "news-02",
          "register": "news",
          "faTokens": 108,
          "enTokens": 34,
          "ratio": 3.176
        },
        {
          "id": "news-03",
          "register": "news",
          "faTokens": 96,
          "enTokens": 24,
          "ratio": 4
        },
        {
          "id": "technical-01",
          "register": "technical",
          "faTokens": 94,
          "enTokens": 29,
          "ratio": 3.241
        },
        {
          "id": "technical-02",
          "register": "technical",
          "faTokens": 77,
          "enTokens": 22,
          "ratio": 3.5
        },
        {
          "id": "technical-03",
          "register": "technical",
          "faTokens": 85,
          "enTokens": 24,
          "ratio": 3.542
        },
        {
          "id": "technical-04",
          "register": "technical",
          "faTokens": 78,
          "enTokens": 25,
          "ratio": 3.12
        },
        {
          "id": "policy-01",
          "register": "policy",
          "faTokens": 88,
          "enTokens": 25,
          "ratio": 3.52
        },
        {
          "id": "policy-02",
          "register": "policy",
          "faTokens": 82,
          "enTokens": 18,
          "ratio": 4.556
        },
        {
          "id": "conversational-01",
          "register": "conversational",
          "faTokens": 73,
          "enTokens": 25,
          "ratio": 2.92
        },
        {
          "id": "conversational-02",
          "register": "conversational",
          "faTokens": 64,
          "enTokens": 22,
          "ratio": 2.909
        },
        {
          "id": "conversational-03",
          "register": "conversational",
          "faTokens": 66,
          "enTokens": 25,
          "ratio": 2.64
        },
        {
          "id": "scientific-01",
          "register": "scientific",
          "faTokens": 100,
          "enTokens": 28,
          "ratio": 3.571
        },
        {
          "id": "scientific-02",
          "register": "scientific",
          "faTokens": 71,
          "enTokens": 25,
          "ratio": 2.84
        },
        {
          "id": "scientific-03",
          "register": "scientific",
          "faTokens": 90,
          "enTokens": 25,
          "ratio": 3.6
        },
        {
          "id": "product-01",
          "register": "product",
          "faTokens": 67,
          "enTokens": 24,
          "ratio": 2.792
        },
        {
          "id": "product-02",
          "register": "product",
          "faTokens": 72,
          "enTokens": 20,
          "ratio": 3.6
        },
        {
          "id": "finance-01",
          "register": "finance",
          "faTokens": 80,
          "enTokens": 24,
          "ratio": 3.333
        },
        {
          "id": "finance-02",
          "register": "finance",
          "faTokens": 77,
          "enTokens": 20,
          "ratio": 3.85
        },
        {
          "id": "education-01",
          "register": "education",
          "faTokens": 85,
          "enTokens": 26,
          "ratio": 3.269
        },
        {
          "id": "education-02",
          "register": "education",
          "faTokens": 61,
          "enTokens": 24,
          "ratio": 2.542
        },
        {
          "id": "legal-01",
          "register": "legal",
          "faTokens": 85,
          "enTokens": 31,
          "ratio": 2.742
        },
        {
          "id": "legal-02",
          "register": "legal",
          "faTokens": 57,
          "enTokens": 21,
          "ratio": 2.714
        },
        {
          "id": "literary-01",
          "register": "literary",
          "faTokens": 83,
          "enTokens": 26,
          "ratio": 3.192
        },
        {
          "id": "literary-02",
          "register": "literary",
          "faTokens": 66,
          "enTokens": 20,
          "ratio": 3.3
        },
        {
          "id": "medical-01",
          "register": "medical",
          "faTokens": 71,
          "enTokens": 19,
          "ratio": 3.737
        },
        {
          "id": "instruction-01",
          "register": "instruction",
          "faTokens": 79,
          "enTokens": 23,
          "ratio": 3.435
        },
        {
          "id": "instruction-02",
          "register": "instruction",
          "faTokens": 57,
          "enTokens": 26,
          "ratio": 2.192
        }
      ],
      "faTokens": 2216,
      "enTokens": 689,
      "ratio": 3.216,
      "faTokensPerChar": 0.723,
      "enTokensPerChar": 0.1918,
      "faCharsPerToken": 1.383,
      "rank": 13,
      "tied": true
    },
    {
      "id": "mistral-v3",
      "label": "Mistral v3",
      "vendor": "Mistral AI",
      "repo": "Xenova/mistral-tokenizer-v3",
      "mirror": true,
      "persianSpecific": false,
      "note": "32k SentencePiece vocabulary.",
      "revision": "b251f946a5b6b16c36963ea8848611bce67dfbf1",
      "fileSha256": "e553af6fff7d7ad76e830608b218c5c0b0822998d5a1a96099a74cd3c1cb1a49",
      "measuredAt": "2026-08-19",
      "vocabSize": 32768,
      "fingerprint": "c557e2b72ac82990",
      "items": [
        {
          "id": "news-01",
          "register": "news",
          "faTokens": 147,
          "enTokens": 34,
          "ratio": 4.324
        },
        {
          "id": "news-02",
          "register": "news",
          "faTokens": 141,
          "enTokens": 36,
          "ratio": 3.917
        },
        {
          "id": "news-03",
          "register": "news",
          "faTokens": 123,
          "enTokens": 27,
          "ratio": 4.556
        },
        {
          "id": "technical-01",
          "register": "technical",
          "faTokens": 126,
          "enTokens": 29,
          "ratio": 4.345
        },
        {
          "id": "technical-02",
          "register": "technical",
          "faTokens": 112,
          "enTokens": 24,
          "ratio": 4.667
        },
        {
          "id": "technical-03",
          "register": "technical",
          "faTokens": 120,
          "enTokens": 25,
          "ratio": 4.8
        },
        {
          "id": "technical-04",
          "register": "technical",
          "faTokens": 109,
          "enTokens": 27,
          "ratio": 4.037
        },
        {
          "id": "policy-01",
          "register": "policy",
          "faTokens": 119,
          "enTokens": 29,
          "ratio": 4.103
        },
        {
          "id": "policy-02",
          "register": "policy",
          "faTokens": 119,
          "enTokens": 20,
          "ratio": 5.95
        },
        {
          "id": "conversational-01",
          "register": "conversational",
          "faTokens": 96,
          "enTokens": 26,
          "ratio": 3.692
        },
        {
          "id": "conversational-02",
          "register": "conversational",
          "faTokens": 79,
          "enTokens": 24,
          "ratio": 3.292
        },
        {
          "id": "conversational-03",
          "register": "conversational",
          "faTokens": 86,
          "enTokens": 25,
          "ratio": 3.44
        },
        {
          "id": "scientific-01",
          "register": "scientific",
          "faTokens": 132,
          "enTokens": 30,
          "ratio": 4.4
        },
        {
          "id": "scientific-02",
          "register": "scientific",
          "faTokens": 109,
          "enTokens": 27,
          "ratio": 4.037
        },
        {
          "id": "scientific-03",
          "register": "scientific",
          "faTokens": 123,
          "enTokens": 26,
          "ratio": 4.731
        },
        {
          "id": "product-01",
          "register": "product",
          "faTokens": 101,
          "enTokens": 24,
          "ratio": 4.208
        },
        {
          "id": "product-02",
          "register": "product",
          "faTokens": 100,
          "enTokens": 22,
          "ratio": 4.545
        },
        {
          "id": "finance-01",
          "register": "finance",
          "faTokens": 105,
          "enTokens": 24,
          "ratio": 4.375
        },
        {
          "id": "finance-02",
          "register": "finance",
          "faTokens": 100,
          "enTokens": 21,
          "ratio": 4.762
        },
        {
          "id": "education-01",
          "register": "education",
          "faTokens": 117,
          "enTokens": 28,
          "ratio": 4.179
        },
        {
          "id": "education-02",
          "register": "education",
          "faTokens": 86,
          "enTokens": 24,
          "ratio": 3.583
        },
        {
          "id": "legal-01",
          "register": "legal",
          "faTokens": 119,
          "enTokens": 33,
          "ratio": 3.606
        },
        {
          "id": "legal-02",
          "register": "legal",
          "faTokens": 82,
          "enTokens": 23,
          "ratio": 3.565
        },
        {
          "id": "literary-01",
          "register": "literary",
          "faTokens": 103,
          "enTokens": 26,
          "ratio": 3.962
        },
        {
          "id": "literary-02",
          "register": "literary",
          "faTokens": 83,
          "enTokens": 20,
          "ratio": 4.15
        },
        {
          "id": "medical-01",
          "register": "medical",
          "faTokens": 98,
          "enTokens": 19,
          "ratio": 5.158
        },
        {
          "id": "instruction-01",
          "register": "instruction",
          "faTokens": 102,
          "enTokens": 23,
          "ratio": 4.435
        },
        {
          "id": "instruction-02",
          "register": "instruction",
          "faTokens": 81,
          "enTokens": 27,
          "ratio": 3
        }
      ],
      "faTokens": 3018,
      "enTokens": 723,
      "ratio": 4.174,
      "faTokensPerChar": 0.9847,
      "enTokensPerChar": 0.2013,
      "faCharsPerToken": 1.016,
      "rank": 16,
      "tied": true
    },
    {
      "id": "maral-7b",
      "label": "Maral 7B",
      "vendor": "MaralGPT",
      "repo": "MaralGPT/Maral-7B-alpha-1",
      "mirror": false,
      "persianSpecific": true,
      "note": "A Persian community model built on Mistral.",
      "revision": "2ab5ca2a0d1a4454a78b4ca911e595bb9da2fe2f",
      "fileSha256": "fc4f0bd70b3709312d9d1d9e5ba674794b6bc5abc17429897a540f93882f25fc",
      "measuredAt": "2026-08-19",
      "vocabSize": 32000,
      "fingerprint": "3e48030f4beccfad",
      "items": [
        {
          "id": "news-01",
          "register": "news",
          "faTokens": 147,
          "enTokens": 34,
          "ratio": 4.324
        },
        {
          "id": "news-02",
          "register": "news",
          "faTokens": 141,
          "enTokens": 36,
          "ratio": 3.917
        },
        {
          "id": "news-03",
          "register": "news",
          "faTokens": 123,
          "enTokens": 27,
          "ratio": 4.556
        },
        {
          "id": "technical-01",
          "register": "technical",
          "faTokens": 126,
          "enTokens": 29,
          "ratio": 4.345
        },
        {
          "id": "technical-02",
          "register": "technical",
          "faTokens": 112,
          "enTokens": 24,
          "ratio": 4.667
        },
        {
          "id": "technical-03",
          "register": "technical",
          "faTokens": 120,
          "enTokens": 25,
          "ratio": 4.8
        },
        {
          "id": "technical-04",
          "register": "technical",
          "faTokens": 109,
          "enTokens": 27,
          "ratio": 4.037
        },
        {
          "id": "policy-01",
          "register": "policy",
          "faTokens": 119,
          "enTokens": 29,
          "ratio": 4.103
        },
        {
          "id": "policy-02",
          "register": "policy",
          "faTokens": 119,
          "enTokens": 20,
          "ratio": 5.95
        },
        {
          "id": "conversational-01",
          "register": "conversational",
          "faTokens": 96,
          "enTokens": 26,
          "ratio": 3.692
        },
        {
          "id": "conversational-02",
          "register": "conversational",
          "faTokens": 79,
          "enTokens": 24,
          "ratio": 3.292
        },
        {
          "id": "conversational-03",
          "register": "conversational",
          "faTokens": 86,
          "enTokens": 25,
          "ratio": 3.44
        },
        {
          "id": "scientific-01",
          "register": "scientific",
          "faTokens": 132,
          "enTokens": 30,
          "ratio": 4.4
        },
        {
          "id": "scientific-02",
          "register": "scientific",
          "faTokens": 109,
          "enTokens": 27,
          "ratio": 4.037
        },
        {
          "id": "scientific-03",
          "register": "scientific",
          "faTokens": 123,
          "enTokens": 26,
          "ratio": 4.731
        },
        {
          "id": "product-01",
          "register": "product",
          "faTokens": 101,
          "enTokens": 24,
          "ratio": 4.208
        },
        {
          "id": "product-02",
          "register": "product",
          "faTokens": 100,
          "enTokens": 22,
          "ratio": 4.545
        },
        {
          "id": "finance-01",
          "register": "finance",
          "faTokens": 105,
          "enTokens": 24,
          "ratio": 4.375
        },
        {
          "id": "finance-02",
          "register": "finance",
          "faTokens": 100,
          "enTokens": 21,
          "ratio": 4.762
        },
        {
          "id": "education-01",
          "register": "education",
          "faTokens": 117,
          "enTokens": 28,
          "ratio": 4.179
        },
        {
          "id": "education-02",
          "register": "education",
          "faTokens": 86,
          "enTokens": 24,
          "ratio": 3.583
        },
        {
          "id": "legal-01",
          "register": "legal",
          "faTokens": 119,
          "enTokens": 33,
          "ratio": 3.606
        },
        {
          "id": "legal-02",
          "register": "legal",
          "faTokens": 82,
          "enTokens": 23,
          "ratio": 3.565
        },
        {
          "id": "literary-01",
          "register": "literary",
          "faTokens": 103,
          "enTokens": 26,
          "ratio": 3.962
        },
        {
          "id": "literary-02",
          "register": "literary",
          "faTokens": 83,
          "enTokens": 20,
          "ratio": 4.15
        },
        {
          "id": "medical-01",
          "register": "medical",
          "faTokens": 98,
          "enTokens": 19,
          "ratio": 5.158
        },
        {
          "id": "instruction-01",
          "register": "instruction",
          "faTokens": 102,
          "enTokens": 23,
          "ratio": 4.435
        },
        {
          "id": "instruction-02",
          "register": "instruction",
          "faTokens": 81,
          "enTokens": 27,
          "ratio": 3
        }
      ],
      "faTokens": 3018,
      "enTokens": 723,
      "ratio": 4.174,
      "faTokensPerChar": 0.9847,
      "enTokensPerChar": 0.2013,
      "faCharsPerToken": 1.016,
      "rank": 16,
      "tied": true
    },
    {
      "id": "claude",
      "label": "Claude",
      "vendor": "Anthropic",
      "repo": "Xenova/claude-tokenizer",
      "mirror": true,
      "persianSpecific": false,
      "note": "Anthropic publishes no tokenizer file; this is the community mirror in common use.",
      "revision": "cae688821ea05490de49a6d3faa36468a4672fad",
      "fileSha256": "c241737df24b4e7f7c9af4fdcee29a0ca903dcb288a8b753bc346a3092911767",
      "measuredAt": "2026-08-19",
      "vocabSize": 65000,
      "fingerprint": "4773fa1b568b99b8",
      "items": [
        {
          "id": "news-01",
          "register": "news",
          "faTokens": 145,
          "enTokens": 34,
          "ratio": 4.265
        },
        {
          "id": "news-02",
          "register": "news",
          "faTokens": 146,
          "enTokens": 34,
          "ratio": 4.294
        },
        {
          "id": "news-03",
          "register": "news",
          "faTokens": 127,
          "enTokens": 25,
          "ratio": 5.08
        },
        {
          "id": "technical-01",
          "register": "technical",
          "faTokens": 129,
          "enTokens": 29,
          "ratio": 4.448
        },
        {
          "id": "technical-02",
          "register": "technical",
          "faTokens": 114,
          "enTokens": 23,
          "ratio": 4.957
        },
        {
          "id": "technical-03",
          "register": "technical",
          "faTokens": 120,
          "enTokens": 24,
          "ratio": 5
        },
        {
          "id": "technical-04",
          "register": "technical",
          "faTokens": 114,
          "enTokens": 26,
          "ratio": 4.385
        },
        {
          "id": "policy-01",
          "register": "policy",
          "faTokens": 121,
          "enTokens": 26,
          "ratio": 4.654
        },
        {
          "id": "policy-02",
          "register": "policy",
          "faTokens": 119,
          "enTokens": 18,
          "ratio": 6.611
        },
        {
          "id": "conversational-01",
          "register": "conversational",
          "faTokens": 95,
          "enTokens": 25,
          "ratio": 3.8
        },
        {
          "id": "conversational-02",
          "register": "conversational",
          "faTokens": 80,
          "enTokens": 22,
          "ratio": 3.636
        },
        {
          "id": "conversational-03",
          "register": "conversational",
          "faTokens": 86,
          "enTokens": 25,
          "ratio": 3.44
        },
        {
          "id": "scientific-01",
          "register": "scientific",
          "faTokens": 138,
          "enTokens": 30,
          "ratio": 4.6
        },
        {
          "id": "scientific-02",
          "register": "scientific",
          "faTokens": 105,
          "enTokens": 26,
          "ratio": 4.038
        },
        {
          "id": "scientific-03",
          "register": "scientific",
          "faTokens": 126,
          "enTokens": 25,
          "ratio": 5.04
        },
        {
          "id": "product-01",
          "register": "product",
          "faTokens": 101,
          "enTokens": 24,
          "ratio": 4.208
        },
        {
          "id": "product-02",
          "register": "product",
          "faTokens": 106,
          "enTokens": 21,
          "ratio": 5.048
        },
        {
          "id": "finance-01",
          "register": "finance",
          "faTokens": 104,
          "enTokens": 24,
          "ratio": 4.333
        },
        {
          "id": "finance-02",
          "register": "finance",
          "faTokens": 103,
          "enTokens": 20,
          "ratio": 5.15
        },
        {
          "id": "education-01",
          "register": "education",
          "faTokens": 115,
          "enTokens": 26,
          "ratio": 4.423
        },
        {
          "id": "education-02",
          "register": "education",
          "faTokens": 82,
          "enTokens": 24,
          "ratio": 3.417
        },
        {
          "id": "legal-01",
          "register": "legal",
          "faTokens": 120,
          "enTokens": 31,
          "ratio": 3.871
        },
        {
          "id": "legal-02",
          "register": "legal",
          "faTokens": 79,
          "enTokens": 21,
          "ratio": 3.762
        },
        {
          "id": "literary-01",
          "register": "literary",
          "faTokens": 108,
          "enTokens": 26,
          "ratio": 4.154
        },
        {
          "id": "literary-02",
          "register": "literary",
          "faTokens": 87,
          "enTokens": 20,
          "ratio": 4.35
        },
        {
          "id": "medical-01",
          "register": "medical",
          "faTokens": 96,
          "enTokens": 19,
          "ratio": 5.053
        },
        {
          "id": "instruction-01",
          "register": "instruction",
          "faTokens": 104,
          "enTokens": 23,
          "ratio": 4.522
        },
        {
          "id": "instruction-02",
          "register": "instruction",
          "faTokens": 83,
          "enTokens": 27,
          "ratio": 3.074
        }
      ],
      "faTokens": 3053,
      "enTokens": 698,
      "ratio": 4.374,
      "faTokensPerChar": 0.9961,
      "enTokensPerChar": 0.1943,
      "faCharsPerToken": 1.004,
      "rank": 18,
      "tied": false
    },
    {
      "id": "grok-1",
      "label": "Grok-1",
      "vendor": "xAI",
      "repo": "Xenova/grok-1-tokenizer",
      "mirror": true,
      "persianSpecific": false,
      "note": "",
      "revision": "40ee9ae4aed428f623d105e0d85afbbb7985c4b6",
      "fileSha256": "f9ea0625debc3e7ee0f0b3c00e933e230ad9c88afe16369f865e8fde3d836faa",
      "measuredAt": "2026-08-19",
      "vocabSize": 131072,
      "fingerprint": "d6ed57b72a18068b",
      "items": [
        {
          "id": "news-01",
          "register": "news",
          "faTokens": 141,
          "enTokens": 34,
          "ratio": 4.147
        },
        {
          "id": "news-02",
          "register": "news",
          "faTokens": 146,
          "enTokens": 35,
          "ratio": 4.171
        },
        {
          "id": "news-03",
          "register": "news",
          "faTokens": 128,
          "enTokens": 24,
          "ratio": 5.333
        },
        {
          "id": "technical-01",
          "register": "technical",
          "faTokens": 128,
          "enTokens": 28,
          "ratio": 4.571
        },
        {
          "id": "technical-02",
          "register": "technical",
          "faTokens": 115,
          "enTokens": 22,
          "ratio": 5.227
        },
        {
          "id": "technical-03",
          "register": "technical",
          "faTokens": 114,
          "enTokens": 24,
          "ratio": 4.75
        },
        {
          "id": "technical-04",
          "register": "technical",
          "faTokens": 119,
          "enTokens": 26,
          "ratio": 4.577
        },
        {
          "id": "policy-01",
          "register": "policy",
          "faTokens": 130,
          "enTokens": 23,
          "ratio": 5.652
        },
        {
          "id": "policy-02",
          "register": "policy",
          "faTokens": 113,
          "enTokens": 18,
          "ratio": 6.278
        },
        {
          "id": "conversational-01",
          "register": "conversational",
          "faTokens": 96,
          "enTokens": 25,
          "ratio": 3.84
        },
        {
          "id": "conversational-02",
          "register": "conversational",
          "faTokens": 81,
          "enTokens": 22,
          "ratio": 3.682
        },
        {
          "id": "conversational-03",
          "register": "conversational",
          "faTokens": 86,
          "enTokens": 25,
          "ratio": 3.44
        },
        {
          "id": "scientific-01",
          "register": "scientific",
          "faTokens": 140,
          "enTokens": 30,
          "ratio": 4.667
        },
        {
          "id": "scientific-02",
          "register": "scientific",
          "faTokens": 102,
          "enTokens": 26,
          "ratio": 3.923
        },
        {
          "id": "scientific-03",
          "register": "scientific",
          "faTokens": 128,
          "enTokens": 25,
          "ratio": 5.12
        },
        {
          "id": "product-01",
          "register": "product",
          "faTokens": 90,
          "enTokens": 24,
          "ratio": 3.75
        },
        {
          "id": "product-02",
          "register": "product",
          "faTokens": 105,
          "enTokens": 20,
          "ratio": 5.25
        },
        {
          "id": "finance-01",
          "register": "finance",
          "faTokens": 110,
          "enTokens": 24,
          "ratio": 4.583
        },
        {
          "id": "finance-02",
          "register": "finance",
          "faTokens": 104,
          "enTokens": 20,
          "ratio": 5.2
        },
        {
          "id": "education-01",
          "register": "education",
          "faTokens": 110,
          "enTokens": 26,
          "ratio": 4.231
        },
        {
          "id": "education-02",
          "register": "education",
          "faTokens": 80,
          "enTokens": 24,
          "ratio": 3.333
        },
        {
          "id": "legal-01",
          "register": "legal",
          "faTokens": 120,
          "enTokens": 32,
          "ratio": 3.75
        },
        {
          "id": "legal-02",
          "register": "legal",
          "faTokens": 77,
          "enTokens": 21,
          "ratio": 3.667
        },
        {
          "id": "literary-01",
          "register": "literary",
          "faTokens": 116,
          "enTokens": 26,
          "ratio": 4.462
        },
        {
          "id": "literary-02",
          "register": "literary",
          "faTokens": 95,
          "enTokens": 20,
          "ratio": 4.75
        },
        {
          "id": "medical-01",
          "register": "medical",
          "faTokens": 96,
          "enTokens": 19,
          "ratio": 5.053
        },
        {
          "id": "instruction-01",
          "register": "instruction",
          "faTokens": 105,
          "enTokens": 23,
          "ratio": 4.565
        },
        {
          "id": "instruction-02",
          "register": "instruction",
          "faTokens": 79,
          "enTokens": 26,
          "ratio": 3.038
        }
      ],
      "faTokens": 3054,
      "enTokens": 692,
      "ratio": 4.413,
      "faTokensPerChar": 0.9964,
      "enTokensPerChar": 0.1927,
      "faCharsPerToken": 1.004,
      "rank": 19,
      "tied": false
    }
  ],
  "skipped": []
}