Qwen: Qwen3.8 2.4t A95b
Qwen's largest mixture-of-experts model with 2.4 trillion parameters, 95B of them active. Always reasons before answering, which suits hard analysis, mathematics and coding. Text only, with tool calling and a 256k context.
- Maker
- Qwen
- Model ID
qwen/qwen3.8-2.4t-a95b- Context length
- 262,144 tokens
- Input
- text
- Output
- text
- Capabilities
- tool calling, structured output, and reasoning
Providers
- Lyceum: Data centers: France, Finland, and Germany; Zero data retention
- TensorX: Data centers: Ireland and Finland; Zero data retention
Facts as Markdown · Quickstart
More models by Qwen
- Qwen: Qwen3 235b A22b 2507 Qwen · 250,000 tokens · text
- Qwen: Qwen3 Coder Next Qwen · 262,144 tokens · text
- Qwen: Qwen3 Embedding 8B Qwen · 32,768 tokens · embeddings
- Qwen: Qwen3 VL Reranker 8B Qwen · 32,768 tokens · rerank
- Qwen: Qwen3.5 2B Qwen · 32,768 tokens · decisions
- Qwen: Qwen3.5 397b A17b Qwen · 262,144 tokens · text
More Text models
- DeepSeek: DeepSeek V4 Flash 0731 DeepSeek · 256,000 tokens · text
- DeepSeek: DeepSeek V4 PRO 0813 DeepSeek · 1,048,576 tokens · text
- DeepSeek: DeepSeek V4.1 Flash DeepSeek · 1,048,576 tokens · text
- DeepSeek: DeepSeek V4.1 Flash 1017 DeepSeek · 1,048,576 tokens · text
- Google: Gemma 4 26B A4B IT Google · 256,000 tokens · text
- Google: Gemma 4 31B IT Google · 262,144 tokens · text