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...

Pricing and specs

Providerqwen
Input price$0.230 / 1M tokens
Output price$2.30 / 1M tokens
Context window262k tokens
Max outputNot specified
Modalitiestext
Model IDqwen/qwen3-235b-a22b-thinking-2507

What a typical call costs

A 1,000-token prompt with a 500-token answer on Qwen: Qwen3 235B A22B Thinking 2507 costs roughly $0.00138. Use the OpenRouter cost calculator to price your own prompt across 400+ models, or the token counter to measure prompt length first.

Run your prompts on Qwen: Qwen3 235B A22B Thinking 2507

PromNest stores your prompt library and lets you run any prompt on Qwen: Qwen3 235B A22B Thinking 2507 with your own OpenRouter key — including side-by-side comparison with other models. Browse 1,000+ ready-made prompt templates to start.

Alternatives from qwen

Qwen: Qwen Plus 0728Qwen: Qwen Plus 0728 (thinking)Qwen: Qwen-PlusQwen: Qwen2.5 7B InstructQwen: Qwen2.5 VL 72B InstructQwen: Qwen3 14BQwen: Qwen3 235B A22BQwen: Qwen3 235B A22B Instruct 2507

← Back to the full AI model catalog