qwen
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...
Input Cost
$0.15
per 1M tokens
Output Cost
$1.50
per 1M tokens
Context Window
131,072
tokens
Developer ID: qwen/qwen3-235b-a22b-thinking-2507
Related Models
qwen
$0.12/1M
Qwen: Qwen3 VL 8B Thinking
Qwen3-VL-8B-Thinking is the reasoning-optimized variant of the Qwen3-VL-8B multimodal mode...
qwen
$0.78/1M
Qwen: Qwen3 Max Thinking
Qwen3-Max-Thinking is the flagship reasoning model in the Qwen3 series, designed for high-...
qwen
$0.20/1M
Qwen: Qwen3 Coder Flash
Qwen3 Coder Flash is Alibaba's fast and cost efficient version of their proprietary Qwen3 ...