Qwen · OpenRouter
Qwen3 235B A22B Thinking 2507 pricing
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...
OpenRouterOpenRouter API
This is OpenRouter route pricing from the public models API, not a claim about the creator's direct API price.
Tool useReasoningHugging Face
2,000 input + 500 output tokens$0.00265 / request
Calculate a complete workload →