AI Desk · Model Identity · Compare
Anthropic: Claude Fable 5.1 · Anthropic: Claude Opus 5
Identity similarity: highNear-exact agreement · 49/49Agreement this close usually means one tokenizer behind two serving templates, or a small vocabulary revision inside one family. The differing rows below say which strings part.
Anthropic: Claude Fable 5.1 recordAnthropic: Claude Opus 5 recordOpen in the interactive bench
Tokenizer fingerprint
49 of 49 mutually clean strings agree| Test string | Anthropic: Claude Fable 5.1 | Anthropic: Claude Opus 5 | Verdict |
|---|---|---|---|
| en-prose | 24 | 24 | match |
| en-long | 28 | 28 | match |
| spaces-20 | 4 | 4 | match |
| spaces-60 | 6 | 6 | match |
| tabs-20 | 4 | 4 | match |
| newlines-20 | 4 | 4 | match |
| mixed-ws | 9 | 9 | match |
| digits-9 | 3 | 3 | match |
| digits-12 | 4 | 4 | match |
| digits-30 | 10 | 10 | match |
| digits-sep | 7 | 7 | match |
| float-long | 9 | 9 | match |
| zh-common | 18 | 18 | match |
| zh-long | 24 | 24 | match |
| zh-rare | 20 | 20 | match |
| ja-kana | 12 | 12 | match |
| ja-kanji | 16 | 16 | match |
| ko | 19 | 19 | match |
| ru | 14 | 14 | match |
| ar | 21 | 21 | match |
| he | 18 | 18 | match |
| hi | 23 | 23 | match |
| th | 25 | 25 | match |
| el | 22 | 22 | match |
| emoji-basic | 10 | 10 | match |
| emoji-skin | 25 | 25 | match |
| emoji-zwj-family | 14 | 14 | match |
| emoji-zwj-x3 | 42 | 42 | match |
| emoji-flags | 24 | 24 | match |
| emoji-prof | 27 | 27 | match |
| math | 46 | 46 | match |
| boxdraw | 28 | 28 | match |
| combining | 6 | 6 | match |
| cjk-ext-b | 16 | 16 | match |
| surrogates | 54 | 54 | match |
| zalgo | 36 | 36 | match |
| rtl-mix | 11 | 11 | match |
| py-code | 30 | 30 | match |
| py-indent | 19 | 19 | match |
| json | 29 | 29 | match |
| html | 18 | 18 | match |
| regex | 63 | 63 | match |
| camel | 16 | 16 | match |
| snake | 14 | 14 | match |
| rare-word-x5 | 55 | 55 | match |
| repeat-tok | 60 | 60 | match |
| base64 | x | x | – |
| hex | 18 | 18 | match |
| url | 26 | 26 | match |
| uuid | 27 | 27 | match |
Values are the marginal prompt-token cost of each string with the model's fixed overhead subtracted. An x marks a row excluded as corrupt for that model.
API surface
The declared serving contracts differ| Field | Anthropic: Claude Fable 5.1 | Anthropic: Claude Opus 5 | Verdict |
|---|---|---|---|
| Context length | 1000000 | 1000000 | ✓ match |
| Max output tokens | 128000 | 128000 | ✓ match |
| Supported parameters | include_reasoning, max_completion_tokens, max_tokens, reasoning, reasoning_effort, response_format, stop, structured_outputs, tools, verbosity | include_reasoning, max_completion_tokens, max_tokens, reasoning, reasoning_effort, response_format, stop, structured_outputs, temperature, tool_choice, tools, verbosity | × differs |
| Defaults | {} | {} | ✓ match |
| Reasoning contract | {"mandatory":true,"supported_efforts":["max","xhigh","high","medium","low"],"default_effort":"high"} | {"mandatory":false,"default_enabled":true,"supported_efforts":["max","xhigh","high","medium","low"],"default_effort":"high"} | × differs |
| Modalities | text+image+file->text | text+image+file->text | ✓ match |
Method: tokenizer fingerprinting, v1, 50 strings. This page is static and dated; the interactive bench compares any two of the 486 catalogue models.