Alibaba Cloud / Qwen Team model product
Qwen3 32B
Qwen3-32B is a large language model from Alibaba's Qwen3 series.
Updated Aug 12, 2026. Default version: Qwen3 32B
LLMBoard score31.2Qwen3 32B
Coverage40%8 benchmark families
Context window131.1KTokens
Official input price$0.70Alibaba API
Qwen3 32B Specifications
Technical details for the model's default version.
- Version
- Qwen3 32B
- Released
- Apr 29, 2025
- Knowledge cutoff
- Apr 1, 2025
- Parameters
- 32.8B
- Context window
- 131.1K
- Max output
- 16.4K
- Inputs
- text
- Outputs
- text
- Open weights
- Yes
- License
- Apache 2.0
Qwen3 32B Capability Profile
This profile uses the model's current scored version. Arena ratings and prices are shown separately.
Qwen3 32B Benchmark Results
Benchmark scores for Qwen3 32B.
| MultiLF | 73.0% | 01 | 2 | 100.0% | C | |
| Arena Hard | 93.8% | 02 | 26 | 96.0% | C | |
| Aider | 50.2% | 04 | 4 | 0.0% | C | |
| BFCL | 70.3% | 06 | 11 | 50.0% | C | |
| CodeForces | 65.9% | 13 | 16 | 20.0% | C | |
| LiveBench | 74.9% | 15 | 38 | 62.2% | C | |
| AIME 2024 | 81.4% | 23 | 53 | 57.7% | C | |
| LiveCodeBench | 65.7% | 26 | 73 | 65.3% | C | |
| AIME 2025 | 72.9% | 82 | 114 | 28.3% | C |
Qwen3 32B Arena Results
Preference and agent-evaluation results for the default version.
| text | overall | 187 | 1340.1 | 3,926 | N/A | |
| text style control | overall | 197 | 1347.1 | 3,926 | N/A |
Qwen3 32B Runtime Performance
Provider-specific output speed and catalog latency for Qwen3 32B. Runtime does not affect the capability score.
| Sambanova | 327.7 tok/s | 1.08 s | 128K | 128K | |
| Novita | 32.43 tok/s | 0.93 s | 128K | 128K | |
| DeepInfra | 26.95 tok/s | 1.19 s | 128K | 128K |
Output Speed is generated output tokens received per second. Catalog latency is reported separately from observed provider TTFT.
Qwen3 32B Pricing
Official vendor API pricing appears first, followed by individual provider offers.
- Official API
- $0.70 input, $2.8 output per 1M
- Official provider
- Alibaba
- Lowest third-party
- From $0.08 input, $0.28 output per 1M via OpenRouter
- Tracked offerings
- 21
| iFlow | qwen3-32b | global | N/A | N/A | 128K | |
| OpenRouter | qwen/qwen3-32b | global | $0.08 | $0.28 | 131.1K | |
| Deep Infra | Qwen/Qwen3-32B | global | $0.08 | $0.28 | 41K | |
| Abacus | Qwen/Qwen3-32B | global | $0.09 | $0.29 | 131.1K | |
| OVHcloud AI Endpoints | qwen3-32b | global | $0.09 | $0.25 | 32.8K | |
| Cortecs | qwen3-32b | global | $0.099 | $0.299 | 40K | |
| Nebius Token Factory | Qwen/Qwen3-32B | global | $0.10 | $0.30 | 128K | |
| NanoGPT | qwen/qwen3-32b | global | $0.10 | $0.30 | 41K | |
| LLM Gateway | qwen3-32b | global | $0.10 | $0.30 | 41K | |
| NovitaAI | qwen/qwen3-32b-fp8 | global | $0.10 | $0.45 | 41K | |
| Jiekou.AI | qwen/qwen3-32b-fp8 | global | $0.10 | $0.45 | 41K | |
| Kilo Gateway | qwen/qwen3-32b | global | $0.104 | $0.416 | 41K | |
| Merge Gateway | qwen/qwen3-32b | global | $0.15 | $0.60 | 131.1K | |
| Vercel AI Gateway | alibaba/qwen-3-32b | global | $0.16 | $0.64 | 128K | |
| DigitalOcean | alibaba-qwen3-32b | global | $0.25 | $0.55 | 32.8K | |
| Alibaba (China) | qwen3-32b | global | $0.287 | $1.15 | 131.1K | |
| Helicone | qwen3-32b | global | $0.29 | $0.59 | 131.1K | |
| Hugging Face | Qwen/Qwen3-32B | global | $0.29 | $0.59 | 131.1K | |
| Alibaba | qwen3-32b | global | $0.70 | $2.8 | 131.1K | |
| Pioneer | Qwen/Qwen3-32B | global | $0.90 | $0.90 | 131.1K |
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
Qwen3 32B Versions
Available versions of this model. The score column identifies the version used in the overall ranking.
| Qwen3 32B | 31.2 | 32.8B | 131.1K | 16.4K | Yes | Apache 2.0 |
What is Qwen3 32B?
Key information about Qwen3 32B and its available data.
Qwen3-32B is a large language model from Alibaba's Qwen3 series. 8 billion parameters, a 128k token context window, support for 119 languages, and hybrid thinking modes allowing switching between deep reasoning and fast responses.
It demonstrates strong performance in reasoning, instruction-following, and agent capabilities.
Data as of 2026-08-11.
Qwen3 32B vs nearby models
Open a comparison with the three ranked models immediately above and below this model.
Models similar to Qwen3 32B
Recommendations prioritize the same model type and family, then the closest LLMBoard score.
AC
Qwen3 VL 30B A3B Thinking
Alibaba Cloud / Qwen Team
AC
Qwen3 30B A3B
Alibaba Cloud / Qwen Team
AC
Qwen3.5 4B
Alibaba Cloud / Qwen Team
AC
Qwen3 VL 30B A3B
Alibaba Cloud / Qwen Team
AC
Qwen3 VL 8B Thinking
Alibaba Cloud / Qwen Team
AC
Qwen3 VL 32B
Alibaba Cloud / Qwen Team
FAQ
Common questions about Qwen3 32B.
When was Qwen3 32B released?
Qwen3 32B's default version was released on Apr 29, 2025.
How much does Qwen3 32B cost?
Qwen3 32B's official API price is $0.70 per million input tokens and $2.8 per million output tokens via Alibaba. The lowest tracked third-party offer starts at $0.08 input and $0.28 output via OpenRouter.
Who created Qwen3 32B?
Qwen3 32B was created by Alibaba Cloud / Qwen Team.
What is the context window for Qwen3 32B?
The default version has a 131.1K token context window.
Is Qwen3 32B open weight?
Yes. The default version is marked as open weight under Apache 2.0.
How many API providers offer Qwen3 32B?
21 provider offerings are linked to the default version.
What models should I compare Qwen3 32B with?
Nearby ranked alternatives include Qwen3 30B A3B, GPT-4.1, Qwen3 VL 30B A3B Thinking.