qwen-qwq-32b
Qwen QwQ-32B is a chat model developed by Qwen, capable of handling a wide range of tasks including streaming, code execution, and function calling. It is genuinely best at extended thinking and generating structured output, making it a versatile tool for various applications.
Input
Output
Context
131K
Max Output
33K
Parameters
32B
Technical Specifications
Model TypeChat
Context Window131,072 tokens
Max Output Tokens32,768 tokens
Parameters32B
Release DateMar 5, 2025
Training CutoffNov 28, 2024
Licenseapache-2.0
Open SourceYes
Input Modalities
Text
Output Modalities
Text
Capabilities
Benchmarks
Artificial Analysis
8.2%
HLE
25.0%
AA-LCR
78.0%
AIME
59.3%
GPQA
29.0%
AIME 2025
38.8%
IFBench
35.8%
SciCode
95.7%
MATH 500
76.4%
MMLU Pro
29.0
Math
63.1%
LiveCodeBench
13.4
Intelligence
Resources & Links
Estimated GPU Requirements for qwen-qwq-32b
Estimates based on INT8 quantization. Actual requirements vary by framework and configuration.
More Models by Qwen
qwen3-235b-a22b-thinking-2507262K context · 235B · Open weightsQwen 3 4B Thinking 2507262K context · 4B · Open weightsQwen 2.5 32B Instruct33K context · 32.8B · Open weightsQwen 2.5 14B Instruct AWQ33K context · 14.8B · Open weightsQwen 3 VL 8B Instruct262K context · 8.8B · Open weightsQwen 3.5 9B262K context · 9.7B · Open weights
Related Tools
Data sourced from official provider APIs and documentation
Last updated: Aug 6, 2026
Your AI stack shouldn't stand still.
Every month new models become cheaper, faster, and more capable. Inferbase ensures your application automatically benefits without changing a single API call.