AI Desk · AI Fingerprinting · Compare
AionLabs: Aion-3.0 · Z.ai: GLM 5.3 Flash
Identity similarity: moderatePartial agreement · 42/50Partial agreement is weak lineage evidence on its own. Related vocabularies overlap; so do unrelated ones on easy strings.
AionLabs: Aion-3.0 recordZ.ai: GLM 5.3 Flash recordOpen in the interactive bench
Tokenizer fingerprint
42 of 50 mutually clean strings agree| Test string | AionLabs: Aion-3.0 | Z.ai: GLM 5.3 Flash | Verdict |
|---|---|---|---|
| en-prose | 13 | 14 | differs |
| en-long | 15 | 16 | differs |
| spaces-20 | 3 | 3 | match |
| spaces-60 | 3 | 3 | match |
| tabs-20 | 3 | 3 | match |
| newlines-20 | 4 | 4 | match |
| mixed-ws | 5 | 5 | match |
| digits-9 | 5 | 5 | match |
| digits-12 | 7 | 7 | match |
| digits-30 | 10 | 10 | match |
| digits-sep | 9 | 9 | match |
| float-long | 14 | 14 | match |
| zh-common | 9 | 9 | match |
| zh-long | 10 | 10 | match |
| zh-rare | 18 | 18 | match |
| ja-kana | 10 | 10 | match |
| ja-kanji | 14 | 14 | match |
| ko | 17 | 17 | match |
| ru | 13 | 13 | match |
| ar | 13 | 13 | match |
| he | 24 | 24 | match |
| hi | 27 | 27 | match |
| th | 24 | 24 | match |
| el | 14 | 14 | match |
| emoji-basic | 5 | 5 | match |
| emoji-skin | 10 | 10 | match |
| emoji-zwj-family | 7 | 7 | match |
| emoji-zwj-x3 | 21 | 21 | match |
| emoji-flags | 24 | 24 | match |
| emoji-prof | 15 | 15 | match |
| math | 27 | 27 | match |
| boxdraw | 28 | 28 | match |
| combining | 10 | 10 | match |
| cjk-ext-b | 13 | 13 | match |
| surrogates | 40 | 40 | match |
| zalgo | 22 | 22 | match |
| rtl-mix | 8 | 8 | match |
| py-code | 24 | 25 | differs |
| py-indent | 15 | 15 | match |
| json | 20 | 21 | differs |
| html | 15 | 16 | differs |
| regex | 44 | 44 | match |
| camel | 5 | 5 | match |
| snake | 6 | 6 | match |
| rare-word-x5 | 30 | 31 | differs |
| repeat-tok | 21 | 22 | differs |
| base64 | 34 | 35 | differs |
| hex | 14 | 14 | match |
| url | 17 | 17 | match |
| uuid | 28 | 28 | 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 | AionLabs: Aion-3.0 | Z.ai: GLM 5.3 Flash | Verdict |
|---|---|---|---|
| Context length | 131072 | 1048576 | × differs |
| Max output tokens | 32768 | 943717 | × differs |
| Supported parameters | include_reasoning, max_tokens, reasoning, response_format, temperature, tool_choice, tools, top_p | frequency_penalty, include_reasoning, logit_bias, logprobs, max_tokens, min_p, parallel_tool_calls, presence_penalty, reasoning, reasoning_effort, repetition_penalty, response_format, seed, stop, structured_outputs, temperature, tool_choice, tools, top_k, top_logprobs, top_p | × differs |
| Defaults | {} | {"temperature":1,"top_p":0.95} | × differs |
| Reasoning contract | {"mandatory":true} | {"mandatory":true,"default_enabled":true,"supported_efforts":["max","high","low"],"default_effort":"max"} | × differs |
| Modalities | text->text | text+image+video->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.