Qwen4-Max: price, features and how it is used
Qwen4-Max is Alibaba’s flagship model, announced on 16 November 2026 and available through Alibaba Cloud Model Studio. The context window is 1M tokens. Alibaba does not announce open weights for it, but says that smaller open models from the Qwen4 series will be released soon.
Provider: Alibaba (Qwen) · Released: November 2026
Intelligence, price and speed by effort level
There is no data yet for this model on effort level, intelligence and price per task.
Data on the Intelligence Index, capability indexes, price per task, speed and latency: Artificial Analysis. Prices per 1M tokens are from official vendor data.
Compare Qwen4-Max with another model
Choose a model and see prices, context and results side by side.
How to use Qwen4-Max
Qwen4-Max is used through Alibaba Cloud Model Studio with the model qwen4-max. The supported regions are China (Beijing), Singapore, Germany (Frankfurt), US (Virginia), Japan (Tokyo) and Hong Kong.
- Open Alibaba Cloud Model Studio and choose a region, for example Frankfurt or Singapore. The price depends on the region.
- Set the model to qwen4-max. The OpenAI Chat Completions, OpenAI Responses, Anthropic Messages and DashScope protocols are supported.
- Thinking is on by default (enable_thinking is true). The reasoning_effort parameter accepts low, high and xhigh, and defaults to high.
- In OpenAI Responses the effort is reasoning.effort, and in the Anthropic protocol it is output_config.effort. The thinking_budget is 131,072 by default.
- The context is 1M tokens, with up to 991,808 input and 131,072 output (262,144 in thinking mode).
- For non-English text, keep in mind that non-English text uses more tokens than English, so the same text costs more. Check the quality of answers on your own real examples.
What it is suited for
- Maximum output 131,072 tokens, and 262,144 tokens in thinking mode.
- Alibaba claims that with Work360 it uses 35% fewer tokens than Qwen3.8-Max at the High level.
What to keep in mind
- The price shown is for the Singapore region. For Virginia, Frankfurt, Tokyo, Beijing and Hong Kong it is $1.413 input and $4.24 output per 1M tokens.
- No published numeric results: Alibaba mentions only rankings and unverifiable internal tests (Work360), so the table is empty.
- The reasoning levels are not documented with token, price and time figures.
Frequently asked questions about Qwen4-Max
- How much does Qwen4-Max cost?
- In Singapore it is USD 1.5 per 1M input and USD 4.5 per 1M output tokens. In Beijing, Hong Kong, Frankfurt, Virginia and Tokyo it is USD 1.413 and USD 4.24. Batch File API in Beijing is USD 0.707 and USD 2.12.
- Is Qwen4-Max free?
- Partly. In Singapore there is a free quota of 1M tokens valid for 90 days, counted from the latest date among Model Studio activation, model release and request approval. After that it is paid.
- How is Qwen4-Max used in other languages?
- Write in your own language via the API and check the answers against your own examples. Non-English text uses more tokens, so the request will cost more than in English. The model documentation does not describe language support.
- How is thinking managed in Qwen4-Max?
- With enable_thinking, which is true by default, and with reasoning_effort set to low, high or xhigh. You can also limit thinking_budget, which is 131,072 by default.
More models in AI Benchmark
- Gemini 4 Neon (Google)
- Gemini 4 Carbon (Google)
- GPT-6 Astra (OpenAI)
- GPT-6.1 Sol (OpenAI)
- GPT-6.1 Terra (OpenAI)
- Muse Wonder Full (Meta)
Intelligence Index, token, TTFT and effort are explained in the AI Benchmark glossary. Prices are per 1M tokens and may change, so check the provider’s official page.
Sources: Alibaba Cloud, Qwen4-Max documentation · Alibaba Cloud, Qwen4-Max announcement
Last checked: 18.11.2026
Compare with other models
Set Qwen4-Max: price, features and how it is used to another model by price, context and benchmarks.