Qwen3-Max
Retired Oct 2026Alibaba's largest Qwen3 model, an MoE with over 1T parameters, offered as an instruct variant with a separate thinking mode.
- Context
- 262K tokens
- Max output
- 66K
Capabilities
What Qwen3-Max accepts and can do, as published by Alibaba Cloud.
- Text Accepts and generates natural-language text.
- Reasoning Thinks before it answers, always on or as a switchable mode.
- Function calling Connect to external tools, APIs, and systems.
- Structured output Return responses in structured formats like JSON.
- Web search Search the web for up-to-date information.
- Code execution Write and execute code in a sandboxed environment.
Common questions
What is Qwen3-Max good for?
Hosted assistant and agent work, at the heavy end of the Qwen3 lineup: a mixture of experts with over 1T parameters.
When is Qwen3-Max not a good fit?
Self-hosting, since the weights are closed, and image, video or audio input, none of which it takes. It sits near the top of Alibaba's price range.
Can I self-host Qwen3-Max?
No. Alibaba Cloud does not publish the weights; Qwen3-Max is available only through hosted APIs.
More from Alibaba Cloud
-
Qwen3-Coder-Plus
1M context
-
Qwen3.5-Flash
1M context
From $0.09 / $0.368 /1M in / out at the cheapest provider
-
Qwen3.5-Plus
1M context
From $0.36 / $2.21 /1M in / out at the cheapest provider
-
Qwen3.6-Flash
1M context
From $0.25 / $1.50 /1M in / out at the cheapest provider
-
Qwen3.6-Plus
1M context
From $0.50 / $3.00 /1M in / out at the cheapest provider
-
Qwen3.7-Flash
1M context
From $0.03 / $0.13 /1M in / out at the cheapest provider
-
Qwen3.7-Max
1M context
From $1.25 / $3.75 /1M in / out at the cheapest provider
-
Qwen3.7-Plus
1M context
From $0.276 / $1.101 /1M in / out at the cheapest provider