AI Desk · AI Fingerprinting · Compare
MoonshotAI: Kimi K2 0711 · MoonshotAI: Kimi K2.5
Identity similarity: lowDivergent fingerprints · 38/48These models tokenise differently. The rows that part here are the English prose, whitespace, CJK, non-Latin script and emoji strings. Divergence on high-information strings separates unrelated vocabularies within a handful of samples.
MoonshotAI: Kimi K2 0711 recordMoonshotAI: Kimi K2.5 recordOpen in the interactive bench
Tokenizer fingerprint
38 of 48 mutually clean strings agree| Test string | MoonshotAI: Kimi K2 0711 | MoonshotAI: Kimi K2.5 | Verdict |
|---|---|---|---|
| en-prose | 13 | 14 | differs |
| en-long | 16 | 16 | match |
| spaces-20 | 2 | 3 | differs |
| spaces-60 | 3 | 3 | match |
| tabs-20 | 4 | 4 | match |
| newlines-20 | 4 | 4 | match |
| mixed-ws | 4 | 6 | differs |
| 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 | 9 | 9 | match |
| zh-long | 10 | 10 | match |
| zh-rare | 16 | 18 | differs |
| ja-kana | 11 | 11 | match |
| ja-kanji | 14 | 14 | match |
| ko | 13 | 13 | match |
| ru | 15 | 16 | differs |
| ar | 15 | 15 | match |
| he | 16 | 16 | match |
| hi | 17 | 17 | match |
| th | 21 | 21 | match |
| el | 22 | 22 | match |
| emoji-basic | 10 | 10 | match |
| emoji-skin | 20 | 20 | match |
| emoji-zwj-family | 9 | 11 | differs |
| emoji-zwj-x3 | 33 | 33 | match |
| emoji-flags | 12 | 14 | differs |
| emoji-prof | 20 | 20 | match |
| math | 26 | 26 | match |
| boxdraw | 24 | 24 | match |
| combining | 14 | 14 | match |
| cjk-ext-b | 16 | 16 | match |
| surrogates | 29 | 31 | differs |
| zalgo | 36 | 36 | match |
| rtl-mix | 8 | 8 | match |
| py-code | 25 | 25 | match |
| py-indent | 15 | 15 | match |
| json | 20 | 20 | match |
| html | 16 | 16 | match |
| regex | 43 | x | – |
| camel | 5 | 6 | differs |
| snake | 6 | 6 | match |
| rare-word-x5 | 31 | 31 | match |
| repeat-tok | 40 | 41 | differs |
| base64 | 33 | 33 | match |
| hex | 11 | 11 | match |
| url | 17 | 17 | match |
| uuid | 25 | x | – |
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 | MoonshotAI: Kimi K2 0711 | MoonshotAI: Kimi K2.5 | Verdict |
|---|---|---|---|
| Context length | 131072 | 262144 | × differs |
| Max output tokens | 98304 | – | × differs |
| Supported parameters | frequency_penalty, max_tokens, presence_penalty, repetition_penalty, seed, stop, temperature, tool_choice, tools, top_k, top_p | frequency_penalty, include_reasoning, logit_bias, logprobs, max_tokens, min_p, presence_penalty, reasoning, repetition_penalty, response_format, seed, stop, structured_outputs, temperature, tool_choice, tools, 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":false,"default_enabled":true} | × differs |
| Modalities | text->text | text+image->text | × differs |
MoonshotAI: Kimi K2.5: OpenRouter's listing shows 235,929 output tokens, which is 90% of the context window, the figure it carries where the provider declares no maximum.
Method: tokenizer fingerprinting, v1, 50 strings. This page is static and dated; the interactive bench compares any two of the 337 measured models.