Alibaba prices Qwen4-Max at $1.50 per million tokens
Alibaba set its flagship model's rate at $1.50 per million input tokens and $4.50 per million output tokens in Singapore. The model page carries no benchmark results.
Developers can now put a price on Qwen4-Max: $1.50 per million tokens sent to the model and $4.50 per million it generates, in Alibaba Cloud's Singapore region. The model page, stamped Nov. 16, supplies the rate that was missing when the model opened in preview for subscribers only. A token is the unit a model uses to count text.
The other five regions, Beijing, Hong Kong, Frankfurt, Virginia and Tokyo, charge $1.413 for input and $4.24 for output. Virginia and Frankfurt are options for anyone who needs to process data outside China. A request with 100,000 input tokens and 10,000 output tokens would run about $0.20 in Singapore. Only Singapore comes with a free allowance: 1 million tokens, valid for 90 days.
What the page lists
- Context: 1,000,000 tokens. Maximum input is 991,808 without reasoning and 983,616 with it.
- Output: up to 131,072 tokens; reasoning and the final answer together are capped at 262,144.
- Inputs and output: text, images and video go in; only text comes out.
- Limits in the global regions: 30,000 requests and 5 million tokens per minute.
- Not supported: batch inference, fine-tuning and self-hosted deployment.
The preview listed a maximum output of 393,216 tokens and `xhigh` as the default reasoning level. The current page sets `high` as the default, with `low`, `high` and `xhigh` available.
No independent testing
The page publishes no benchmark results, so any performance figure is the company's own and no one has replicated it. It gives no general release date and doesn't say whether the subscription preview still runs through Nov. 25. For ordinary users in Mexico nothing changes: the rate matters to developers and companies building Qwen into their products.
Keep reading
- Artificial intelligence
Alibaba opens Qwen4-Max to subscribers only, through Nov. 25
- Artificial intelligence
Google's Gemini 4 Hydrogen is rolling into Search and the app
- Artificial intelligence
Google launches a Gemini API that answers in probabilities
- Artificial intelligence
Four nonprofits sue to halt the most advanced AI



