Qwen3 235B A22B Thinking 2507
ChatAlibaba
qwen/qwen3-235b-a22b-thinking-2507Qwen3-235B-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...
18
credits / gen
About this model
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...
Technical Specifications
Provider
Alibaba
Type
Chat
Context Window
262,144 tokens
Pricing
18 credits
Knowledge Cutoff
2025-06-30
Supported Languages
en, zh
Capabilities
File Support
Can read PDF, DOCX, XLSX & more
Reasoning
Chain-of-thought reasoning exposed
262K Context
Large context window for long documents