Qwen3-Omni Flash
Qwen3-Omni Flash is a model by Alibaba for speech, and one seller in the catalogue publishes a price for it.
text + image + audio + video → text + audio · made by Alibaba
$0.43→$1.66per Mtok in / out
Alibaba
Sold by 1 way
| Seller | Lane | Rate |
|---|---|---|
| Alibaba aggregator | standard | $0.43per Mtok in$1.66per Mtok outmodels.dev · read 2026-08-27 |
About
Qwen3-Omni Flash — a text and audio model from Alibaba, sold by one company from $0.43 in and $1.66 out per million tokens.
It takes text, images, audio and video and returns text and audio, with a context window of 65,536 tokens. It was published in September 2025, trained on material up to April 2024. Its sellers say it can reason step by step and call a tool. The catalogue files it under speak. Only Alibaba sells it, at $0.43 in and $1.66 out per million tokens.
Every current figure
- Maker
- Alibaba
- Register
- model
- Takes
- text + image + audio + video
- Returns
- text + audio
- Context
- 65,536 tokens
- Longest answer
- 16,384 tokens
- Published
- September 2025
- Knowledge to
- April 2024
- Licence
- not read
- Sellers
- 1
- Price
- $0.43 in and $1.66 out per million tokens — Alibaba