Qwen3.8 Flash Next
Open-weight experimental preview of the Qwen4 architecture: hybrid-attention MoE (125B total, 6B active) with vision encoder for coding, agent tasks, and image and video understanding
alibaba/qwen3.8-flash-next- Lab
- Alibaba
- Family
- qwen
- Providers
- 6
- Context
- 262,144
- Output limit
- 131,072
- Knowledge
- -
- Release
- 2026-08-27
- Updated
- 2026-08-27
- Weights
- Open
- Input
- Output types
- Capabilities
- tools, reasoning, temperature
Providers
6| AMD | Alibaba | Qwen3.8-Flash-Next | 262,144 | 131,072 | $0.00 / $0.00 | Yes | Yes | Yes | Yes |
|---|---|---|---|---|---|---|---|---|---|
| Cortecs | Alibaba | qwen3.8-flash-next | 262,144 | 64,000 | $0.201 / $0.50 | Yes | Yes | Yes | Yes |
| Requesty | Alibaba | qwen3.8-flash-next | 262,144 | 262,144 | $0.20 / $0.50 | Yes | Yes | Yes | Yes |
| Requesty | Alibaba | qwen3.8-flash-next@eu | 262,144 | 262,144 | $0.20 / $0.50 | Yes | Yes | Yes | Yes |
| TensorX | Alibaba | qwen/qwen3.8-flash-next | 262,144 | 64,000 | $0.20 / $0.50 | Yes | Yes | Yes | Yes |
| Vultr | Alibaba | qwen3.8-flash-next | 262,144 | 131,072 | $0.10 / $0.20 | Yes | Yes | Yes | Yes |
| No rows match the current search. | |||||||||