Qwen: Qwen3.5-122B-A10B is Alibaba (Qwen)'s model. Its measured tokenizer sits with the Qwen group, which is evidence that it builds on the Qwen tokenizer. Tokenizer reuse is normal engineering, open vocabularies travel between labs, and this says nothing against Alibaba (Qwen)'s authorship of the model itself.
| Signal | Result | Reference |
|---|---|---|
| Tokenizer signature | Measured with 2 corrupt rows excluded | – |
| API surface | No other entry shares this exact contract | api_587e2c6c |
| Context and output | 262,144 context · 262,144 max output | declared |
| Reasoning contract | Optional | declared |
| Serving providers observed | Novita | measured 21 August 2026 |
Click a row to open the full comparison.
An x marks a row excluded as corrupt (caching or provider interference detected during measurement). Overhead subtracted: 10 prompt tokens.
The Qwen3.5 122B-A10B native vision-language model is built on a hybrid architecture that integrates a linear attention mechanism with a sparse mixture-of-experts model, achieving higher inference efficiency. In terms of...
Supported parameters
frequency_penaltyinclude_reasoninglogit_biaslogprobsmax_tokensmin_ppresence_penaltyreasoningrepetition_penaltyresponse_formatseedstopstructured_outputstemperaturetool_choicetoolstop_ktop_logprobstop_p
defaults: {"temperature":0.6,"top_p":0.95,"top_k":20,"frequency_penalty":null,"presence_penalty":null,"repetition_penalty":null}
Cite this page as the evidence record for Qwen: Qwen3.5-122B-A10B: the URL is stable, measurements are dated, and revisions append to the history above. Method: tokenizer fingerprinting, v1, 50 strings. The raw data behind every figure is on the data page.
