AlibabaLanguage modelAvailable
Qwen4-Max
Qwen4-Max has no overall index value yet because there are too few independent benchmark results.
n/aOverall indexhigher is better
n/ano measurement yet
Input $1.413 · Output $4.24
$2.12per 1M tokenslower is better
1Mtokens at once≈ 1,500 A4 pages
Summary
It is fairly expensive: $2.12 per 1M tokens (input and output blended); a typical model costs $1.50.
Qwen4-Max processes text, images, and video and produces text. Alibaba released it on November 16, 2026 for the API and QwenCloud Token Plan. Its context window holds one million tokens. Tool calling and structured output support integration into custom applications.
Thinking can be disabled or set to low, high, or xhigh. When thinking is enabled, reasoning and the final answer share the output limit. The separate preview ID has different limits and time-limited access. API prices differ between Singapore and Global deployments, including Frankfurt.
On this page
Prices
1 offer, prices in US dollars per 1M tokens, blended from 3 parts input and 1 part output.
| Provider | Input | Output | Cache read | Cache write | Blended |
|---|---|---|---|---|---|
| Alibabadirect | $1.413 | $4.24 | $0.071 | $1.767 | $2.12 |
Direct price according to the developer, checked Nov 16, 2026. Open the LLM price index
More price tiers and rules
- Cache write, 1 hour
- $2.827 per 1M tokensSource
- API pricing region
- Global (Frankfurt, Virginia, Tokyo, Hong Kong) and Beijing. Prices per million tokens; cache writes are for 5 minutes or 1 hour.Source
- Singapore API pricing
- $1.50 input and $4.50 output per million tokens. Cache reads cost $0.075; writes cost $1.875 for 5 minutes or $3 for 1 hour.Source
No Alibaba subscription includes this model. Through other providers from $1 a month (Command Code Go, with limited credits). 5 plans in Germany.
| Plan | Monthly list price | Note |
|---|---|---|
| Command Code Go | $1 | with limited credit |
| Command Code GOAT | $10 | with limited credit |
| Command Code Pro | $20 | with limited credit |
| Command Code Max 10× | $100 | with limited credit |
| Command Code Max 20× | $200 | with limited credit |
Similar models
More models of the same kind as Qwen4-Max, chosen by model line, developer, size, price and release date.
- Qwen 3.8 Max 0902Alibaba · same developer
- Qwen 3.7 MaxAlibaba · same developer
- Qwen 3.8 2.4T A95BAlibaba · same developer
- Qwen 3.8 27BAlibaba · same developer
- Qwen 3.8 FlashAlibaba · same developer
- Qwen 3.7 FlashAlibaba · same developer
- Qwen4-Max-PreviewAlibaba · same developer
- Qwen3.8-Omni-FlashAlibaba · same developer
Cost for your load
1,000 requests with 2,000 input and 500 output tokens cost $4.95 with Qwen4-Max.
All specifications
Architecture
- Model parameters
- Qwen reports 2.4 trillion backbone parameters plus an additional 328 billion N-gram embedding parameters. It activates 71 billion parameters per token (mixture of experts). Qwen does not state a combined total there.
- Source: Model parameters
Input and output
- Reasoning
- low, high, and xhigh. Defaults are high and a 131,072-token thinking budget. Thinking can be disabled.
- Source: Reasoning
Watermarking and provenance
AI watermarking refers to invisible signals embedded in the output. Visible marks and provenance metadata are listed separately. Unknown means there is insufficient evidence for that particular field.
- AI watermarking
- Unknown
- no source
- Watermarking technique
- Not known
- no source
Licence and access
- Shutdown
- None announced
- no source
API
Further details
- Output limit
- 262,144 tokens for reasoning and the final answer together; 131,072 tokens without thinking.
- Source: Output limit
- API capabilities
- Tool calling, structured output, streaming, and context caching. Supports Chat Completions, Responses, Anthropic Messages, and DashScope.
- Source: API capabilities
- Preview variant
- qwen4-max-preview remains a separate, time-limited ID with different reasoning and output limits. Checkpoint equivalence is not documented.
- Source: Preview variant
Frequently asked questions about Qwen4-Max
How much does Qwen4-Max cost?
Through the API $1.413 per 1M input tokens and $4.24 output. That is fairly expensive.
How large is the context window of Qwen4-Max?
1 M tokens. That is how much text Qwen4-Max reads at once, about 1,500 A4 pages.
Page count as a rule of thumb from Artificial Analysis
Can I run Qwen4-Max locally?
No. Alibaba does not publish the weights.
Stated by Alibaba
- Open weights
- no
- License
- n/a
- Parameters
- n/a
- active per token
- 71B
Where can I use Qwen4-Max?
Not in a Alibaba subscription. Including other providers that makes 5 plans in Germany, plus the API.
Subscription comparison, prices in Germany incl. VAT
At Alibaba
No Alibaba subscription includes this model.
Through other providers
5 plans from 1 provider, from $1 a month (Command Code Go), with limited credits.
Through the API
$1.413 input, $4.24 output per 1M tokens.
Sources and data date
| Type | Evidence and data date |
|---|---|
| Price | Base API prices without caching or batch discounts. Higher context tiers and other rates are listed under Costs. |
| Input | Documented modalities. Retrieved November 16, 2026. |
| Output | Documented modalities. Retrieved November 16, 2026. |
| Additional source | Qwen4-Max-Preview separate access and limits (retrieved November 16, 2026) |
| Additional source | Alibaba Qwen4-Max release announcement (retrieved November 16, 2026) · Qwen4-Max release announcement · Editorial description reviewed November 16, 2026 |
| Additional source | Qwen4-Max architecture and parameter breakdown (retrieved November 16, 2026) |
| Additional source | Alibaba Cloud Model Studio model releases (retrieved November 16, 2026) |