Built for developers making agents on Alibaba Cloud; the million-token context is the draw, but weights are closed and UK availability is unconfirmed.
Kenning's take
What it does
Qwen3.7-Max is the biggest model in Alibaba's Qwen3.7 series, built for agent work such as programming, office tasks and long autonomous runs. It takes text in and gives text out, with a context window of one million tokens and support for function calling, structured outputs and web search. Developers use it through Alibaba Cloud Model Studio's API, and the weights are not public.
- One million token context window
- Up to 131,072 output tokens
- Thinking mode with long chain of thought
- Function calling and structured outputs
- Built-in web search and context caching
Who it suits. Developers and companies building AI agents for coding and office work on Alibaba Cloud.
Key facts
- Category
- AI models
- Part of
- Qwen
- Made by
- Alibaba (Qwen team)
- Based in
- China
- Released
- 20 May 2026
- How to get it
- through the Alibaba Cloud Model Studio API, in regions including Frankfurt, Virginia, Singapore and Beijing; closed weights
- Works on
- API (Alibaba Cloud Model Studio)
- API price
- $1.65 per million input tokens, $4.951 per million output (Frankfurt, Virginia, Tokyo, Hong Kong); $2.50 and $7.50 in Singapore
Reviews and write-ups
- Decrypt 2026-05-20
FAQ
When was Qwen3.7-Max released?
The current version matches the snapshot qwen3.7-max-2026-05-20, released around Alibaba Cloud's summit on 20 May 2026.
Is Qwen3.7-Max open source?
No. Alibaba keeps the Max model's weights closed and offers it through its Model Studio API.
How much does Qwen3.7-Max cost?
In most global regions it costs $1.65 per million input tokens and $4.951 per million output tokens.
Facts from the maker's own pages and the sources listed. Checked 2026-10-11.
Similar tools
MAI-UI
Alibaba Tongyi research models that operate smartphone apps by looking at the screen and tapping
Qwen-Image-2.0
Alibaba's 7-billion-parameter model that both creates and edits images, with strong text rendering at 2K
Qwen-Image-Layered
Open Alibaba model that splits an image into separate transparent layers so each part can be edited alone
Qwen-Image-3.0
Alibaba's hosted image model built for text-heavy layouts, taking prompts up to 4,500 tokens