Qwen3-VL-32B-Thinking
Meet Qwen3-VL — the most powerful vision-language model in the Qwen series to date.
text + image → text · made by Alibaba
$0.2→$1.5per Mtok in / out
SiliconFlow
Sold by 1 way
| Seller | Lane | Rate |
|---|---|---|
| SiliconFlow aggregator | standard | $0.2per Mtok in$1.5per Mtok outmodels.dev · read 2026-10-31 |
About
Qwen3-VL-32B-Thinking — a text model from Alibaba, sold by one company from $0.2 in and $1.5 out per million tokens.
It takes text and images and returns text, with a context window of 262,000 tokens. It was published in October 2025. Its sellers say it can reason step by step and call a tool. The catalogue files it under reasoning. Its weights are published under APACHE-2.0, so you may run it on your own machine, or buy it from whoever serves it cheapest. Only SiliconFlow sells it, at $0.2 in and $1.5 out per million tokens.
Every current figure
- Maker
- Alibaba
- Register
- model
- Takes
- text + image
- Returns
- text
- Context
- 262,000 tokens
- Published
- October 2025
- Parameters
- 32 billion · read from its own name
- Licence
- apache-2.0
- Sellers
- 1
- Maker's own price
- not read
- Price
- $0.2 in and $1.5 out per million tokens — SiliconFlow
Known as 1 name
Qwen/Qwen3-VL-32B-Thinking