This model is 21 months old. A newer version of it has since shipped: Qwen3.8-27B-FP8 (2026-08-14). The details below are still accurate for this model — it just is not what we would recommend today.
Qwen2.5-Coder 32B
Alibaba / Qwen🇨🇳 China
Best open-source code model at 32B size. Matches GPT-4o on coding benchmarks.
Context window131K tokens
Input / 1M tokens$0.1
Output / 1M tokens$0.3
Version History
Qwen2.5-Coder-32B-Instructmajor
Qwen2.5-Coder 32B matches GPT-4o on HumanEval and competitive coding tasks. Best open-source code model at its parameter count.