AI Desk · AI Fingerprinting · Compare
OpenAI: GPT-6.1 Sol Pro · TheDrummer: UnslopNemo 12B
Insufficient evidenceNot comparable yetAt least one of these records is declared unmeasurable (OpenAI: GPT-6.1 Sol Pro: unmeasurable:caching), so no similarity verdict is published; the declared contracts below can still be compared.
OpenAI: GPT-6.1 Sol Pro recordTheDrummer: UnslopNemo 12B recordOpen in the interactive bench
Tokenizer fingerprint
1 of 29 mutually clean strings agree| Test string | OpenAI: GPT-6.1 Sol Pro | TheDrummer: UnslopNemo 12B | Verdict |
|---|---|---|---|
| en-prose | 49 | 14 | differs |
| en-long | 57 | 18 | differs |
| spaces-20 | 9 | 3 | differs |
| spaces-60 | 16 | 3 | differs |
| tabs-20 | 15 | 4 | differs |
| newlines-20 | x | 7 | – |
| mixed-ws | x | 7 | – |
| digits-9 | 11 | 10 | differs |
| digits-12 | 20 | 13 | differs |
| digits-30 | x | 31 | – |
| digits-sep | x | 13 | – |
| float-long | 29 | 23 | differs |
| zh-common | 45 | 17 | differs |
| zh-long | 60 | 20 | differs |
| zh-rare | x | 21 | – |
| ja-kana | x | 12 | – |
| ja-kanji | 46 | 13 | differs |
| ko | 41 | 12 | differs |
| ru | 40 | 12 | differs |
| ar | 40 | 8 | differs |
| he | 38 | 10 | differs |
| hi | 41 | 14 | differs |
| th | 42 | 17 | differs |
| el | 49 | 13 | differs |
| emoji-basic | 19 | 19 | match |
| emoji-skin | 46 | 39 | differs |
| emoji-zwj-family | x | 18 | – |
| emoji-zwj-x3 | x | 56 | – |
| emoji-flags | x | 31 | – |
| emoji-prof | x | 34 | – |
| math | x | 33 | – |
| boxdraw | x | 30 | – |
| combining | x | 14 | – |
| cjk-ext-b | x | 16 | – |
| surrogates | x | 53 | – |
| zalgo | 4 | 36 | differs |
| rtl-mix | 23 | 6 | differs |
| py-code | x | 25 | – |
| py-indent | 53 | 15 | differs |
| json | x | 25 | – |
| html | x | 17 | – |
| regex | x | 50 | – |
| camel | 26 | 6 | differs |
| snake | 26 | 7 | differs |
| rare-word-x5 | 100 | 35 | differs |
| repeat-tok | 69 | 40 | differs |
| base64 | x | 39 | – |
| hex | 40 | 19 | differs |
| url | 59 | 19 | differs |
| uuid | x | 37 | – |
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 | OpenAI: GPT-6.1 Sol Pro | TheDrummer: UnslopNemo 12B | Verdict |
|---|---|---|---|
| Context length | 1050000 | 1024000 | × differs |
| Max output tokens | 128000 | 819200 | × differs |
| Supported parameters | include_reasoning, max_completion_tokens, max_tokens, reasoning, reasoning_effort, response_format, seed, structured_outputs, tool_choice, tools, verbosity | frequency_penalty, logit_bias, logprobs, max_tokens, presence_penalty, repetition_penalty, seed, stop, structured_outputs, temperature, top_k, top_logprobs, top_p | × differs |
| Defaults | {"temperature":null,"top_p":null,"top_k":null,"frequency_penalty":null,"presence_penalty":null,"repetition_penalty":null} | {} | × differs |
| Reasoning contract | {"mandatory":true,"default_enabled":true,"supported_efforts":["max","xhigh","high","medium","low"],"default_effort":"medium"} | – | × differs |
| Modalities | text+image+file->text | text->text | × differs |
Method: tokenizer fingerprinting, v1, 50 strings. This page is static and dated; the interactive bench compares any two of the 337 measured models.