Intelligence Index
0140.0
> 90% OF MODELS
Qwen
Qwen3.8 2.4T A95B
Updated: September 2026
Qwen3.8 2.4T A95B is an open-weight sparse mixture-of-experts model from Qwen and the open-weight variant of [Qwen3.8 Max](/qwen/qwen3.8-max), with 95 billion active parameters out of 2.4 trillion total. It is...
Specifications
- Context
- 1M
- Input Modalities
- Text
- Output Modalities
- Text
- Input
- $2/M
- Output
- $6/M
Capabilities
TEXTWEBCODINGTHINKINGWRITING
Similarly Priced Models
| Model | Provider | Context | Input Price | Output Price |
|---|---|---|---|---|
| Qwen3.8 Max | Qwen | 1M | $2/M | $6/M |
| Qwen3.8 2.4T A95B (batch) | Qwen | 1M | $2/M | $6/M |
| SpaceXAI: Grok 4.6 | xAI | 500K | $2/M | $6/M |
| Grok 4.5 | xAI | 500K | $2/M | $6/M |
| Claude Sonnet 5 | Anthropic | 1M | $2/M | $10/M |
Performance Metrics
Coding Index
0271.9
> 85% OF MODELS
Agentic Index
0350.4
> 89% OF MODELS
Average Response Performance
- Output Speed
- 38.4 tok/s
- Time To First Token
- 2.84s
- Time To First Answer Token
- 54.93s
- End To End Response Time
- 67.95s
DATA SOURCE: Artificial Analysis