qwen/qwen3-235b-a22b-thinking-2507
Qwen3-235B-A22B-Thinking-2507 is a high-performance, open-weight Mixture-of-Experts (MoE) language model optimized for complex reasoning tasks. It activates 22B of its 235B parameters per forward pass and natively supports up to 262,144...
| Provider | In $/M | Out $/M | Context | Uptime |
|---|---|---|---|---|
| WandBbf16 | $0.100 | $0.100 | 262K | 100% |
| Alibaba | $0.149 | $1.50 | 131K | 99.1% |
| DeepInfrafp8 | $0.230 | $2.30 | 262K | — |
| AtlasCloudfp8 | $0.280 | $2.30 | 128K | 99.4% |
| Novitafp8 | $0.300 | $3.00 | 131K | — |
Qwen3 235B A22B Thinking 2507 costs $0.100 per million input tokens and $0.100 per million output tokens via OpenRouter, making it 55th cheapest of 298 paid models.
Qwen3 235B A22B Thinking 2507 scores 29.5 on the Artificial Analysis Intelligence Index, ranking 83rd of 178 benchmarked models, with a GPQA Diamond score of 79%.
Qwen3 235B A22B Thinking 2507 generates around 79 tokens per second with 382ms time-to-first-token (p50), the 95th fastest tracked model.
Qwen3 235B A22B Thinking 2507 supports a 262K-token context window and can output up to 262K tokens. It accepts text input.