A new generation of open-source, non-thinking mode model powered by Qwen3. This version demonstrates superior Chinese text understanding, augmented logical reasoning, and enhanced capabilities in text generation tasks over the previous iteration (Qwen3-235B-A22B-Instruct-2507).
Try NowChinese text understanding
Efficient non-thinking with strong logic
Text generation without reasoning overhead
131,072 tokens
32,768 tokens
$0.15 per 1M tokens
$1.20 per 1M tokens
$15 per 1K calls
$0.19 per 1K calls