Qwen3 235B A22B 2507
Qwen3 235B A22B 2507 (Reasoning)
Intelligence
12.71
#303 of 520 models
Output speed
66.1 tok/s
Measured generation rate
Cost / index task
$0.081
Measured evaluation workload
End-to-end latency
41ms
First answer 33ms
01 / Position
A model in the landscape
Compare this model with peers across intelligence, speed, cost, and individual evaluations.
02 / Evidence
Benchmark results
25 distinct evaluations grouped by topic. Explore cross-model charts or switch to the full results list.
Domain capability
03 / Operation
Speed, tokens & price
Prices are per million tokens; latency and task costs describe the captured evaluation rather than a universal provider guarantee.
Token pricing
- Input / 1M
- $0.230
- Output / 1M
- $2.30
- Cache read / 1M
- —
- Cache write / 1M
- —
- Cache-hit discount
- —
Response & usage
- Output speed
- 66.1 tok/s
- First chunk
- 3ms
- First answer
- 33ms
- Output / task
- 24.9k
- Reasoning / task
- 22.1k
Evaluation workload
- Canonical input
- 153.3M
- Canonical output
- 86.72M
- Answer / task
- 2.8k
- Index time / task
- 6.1m
- Run cost
- $234.72
By prompt type
| Prompt | Output speed | First chunk | First answer | End to end |
|---|---|---|---|---|
| Long | 66.1 tok/s | 3ms | 33ms | 41ms |
04 / Access
License & availability
Recorded license
Apache 2.0
The publisher-linked license text defines permitted use and redistribution. Review it before using the model.
- Weight access
- Open weights
- Classification
- Permissive
- Reasoning
- Yes
Routing and providers
- OpenRouter ID
- qwen/qwen3-235b-a22b-thinking-2507
- Tokenizer
- Qwen3
- Instruct type
- qwen3
- Captured endpoints
- 3
| Provider | Quantization | Context | tok/s |
|---|---|---|---|
| Alibaba | unknown | 131.1k | — |
| Novita | fp8 | 131.1k | — |
| Venice | fp8 | 128k | — |
05 / Audit
Underlying fields
Inspect the captured model and routing records. Artificial Analysis destination links are omitted.