Qwen3.8-Flash-Next
Key facts · verified Oct 30, 2026
Qwen3.8-Flash-Next is an Alibaba model released on 26 Aug 2026. Its API price is $0.15 per million input tokens and $0.47 per million output tokens (Weights free under qwen-community-1.0; Qwen3.8-Flash API $0.15 / $0.47 per 1M tokens (Alibaba Cloud Model Studio International, qwen3.8-flash, a separate API product from the open weights)). Its context window is 262,144 tokens native, extensible to 1M. It is not yet ranked on LMArena's text leaderboard. Source: huggingface.co.
Alibaba’s Qwen3.8-Flash-Next, published 26 August 2026 as the first open-weight release of the architecture that will underpin Qwen4. A 125B mixture-of-experts with 6B active parameters, plus a 51B n-gram embedding table and a 4B multi-token prediction module, across 48 layers of Gated DeltaNet and Qwen Sparse Attention. Context is 262,144 tokens natively and extensible to 1M. Released under the qwen-community-1.0 licence, not Apache 2.0.
$0.15 / $0.47 per million tokens
262,144 tokens native, extensible to 1M
Not rated
Not ranked
Key Features
Full Specifications
$0.15 / $0.47 per million tokens | |
|---|---|
| Identity | |
| Developer | Alibaba |
| Released | Aug 2026 |
| Status | Open-weight preview of the Qwen4 architecture |
| Licence | qwen-community-1.0 - NOT Apache 2.0 |
| Self-hostable | Yes - weights published on Hugging Face and ModelScope |
| Cost | |
| Blended $/1M tokensinput × 0.75 + output × 0.25 | $0.230 / 1M tokens |
| Input price | $0.150 / 1M tokens |
| Output price | $0.470 / 1M tokens |
| Cached input | Not published |
| Batch discount | Not published |
| Free tier | Weights are free to download under the community licence |
| Capacity | |
| Context window | 262,144 tokens native, extensible to 1M |
| Max output | Not published |
| Capability | |
| Vision in | Yes - text, image and video via a vision encoder |
| Audio in | Not documented |
| Function calling | Not documented |
| Structured output | Not documented |
| Extended reasoning | Not documented |
| Web search | Not documented |
| Access | |
| API | Commercial counterpart Qwen3.8-Flash on QwenCloud |
| Chat app | Not documented for this build |
| Fine-tuning | Self-hosted only |