Intelligence Index
0111.5
> 23% OF MODELS
Qwen
Qwen3 32B
Updated: August 2026
Qwen3-32B is a dense 32.8B parameter causal language model from the Qwen3 series, optimized for both complex reasoning and efficient dialogue. It supports seamless switching between a "thinking" mode for...
Specifications
- Context
- 131K
- Input
- $0.08/M
- Output
- $0.28/M
Capabilities
TEXTWEBCODINGTHINKINGWRITING
Similarly Priced Models
| Model | Provider | Context | Input Price | Output Price |
|---|---|---|---|---|
| Gemma 3 27B | Google | 262K | $0.08/M | $0.45/M |
| ByteDance Seed: Seed 1.6 Flash | ByteDance | 262K | $0.075/M | $0.3/M |
| gpt-oss-safeguard-20b | OpenAI | 131K | $0.075/M | $0.3/M |
| Mistral Small 3.2 24B | MistralAI | 256K | $0.075/M | $0.2/M |
| Gemma 4 26B A4B | Google | 262K | $0.07/M | $0.34/M |
Performance Metrics
Coding Index
0215.3
> 17% OF MODELS
Agentic Index
031.8
> 17% OF MODELS
Average Response Performance
- Output Speed
- 92.5 tok/s
- Time To First Token
- 2.47s
- Time To First Answer Token
- 24.09s
- End To End Response Time
- 29.5s
DATA SOURCE: Artificial Analysis