AlibabaQwen Omni
Qwen3.8 Omni Flash
Qwen3.8 Omni Flash accepts four input modalities but returns text. It supports 113 audio languages and dialects, with optional spatial multichannel audio.
Technical details
Thinking is on by default. The model supports custom tools, web search and caching through Chat Completions or Responses.Qwen3.8 Omni Flash
Practical use
Practical guidance. Use it to write an account of what happens in a recording. Keep time references and speaker evidence in the result so a reviewer can return to the relevant source segment.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Access | API |
| Model identifiers | qwen3.8-omni-flash |
| Context and output | One-million-token context; up to 131,072 output tokens.Qwen3.8 Omni Flash |
| Regions | Beijing, Singapore, Hong Kong, Tokyo, Frankfurt and Virginia.Qwen3.8 Omni Flash |
Related models
- Qwen3.8 LiveTranslate Flash Realtime
Alibaba · Qwen LiveTranslate
Live translation model that understands 60 languages and produces speech in 29, with optional image input.
Launch date not confirmedAPI - Qwen3.8 Omni Flash Realtime
Alibaba · Qwen Omni
Real-time model with streaming audio and video input, text and speech output, and tool support.
Launch date not confirmedAPI