JetBrains' open 12B mixture-of-experts coding model with 2.5B parameters active per token under Apache 2.0
Qwen4-Max
New· 4Alibaba's Qwen flagship on general release from 16 November 2026 after a two-week preview, at $1.50 in and $4.50 out per million tokens internationally, with a 1M-token context and reasoning on by default.
TrustList summary
Not yet independently verified. Alibaba Cloud's list of newly released models dates qwen4-max 16 November 2026, and OpenRouter's listing gives the same day; we first gave 15 November and corrected it on 17 November. Prices and context come from Alibaba Cloud's pricing and model pages. We have not found an independent test of the model. We will update this when it can be confirmed, and remove this note.
Qwen4-Max is the model the two-week qwen4-max-preview was trailing. Where the preview could only be reached through a Token Plan subscription, Alibaba Cloud now lists qwen4-max first on its Model Studio model page, described for complex reasoning and coding, with ordinary per-token prices: $1.50 per million input tokens and $4.50 per million output tokens on the international deployment, the same in thinking and non-thinking modes.
Prices by region and cache
The pricing page quotes one tier for inputs of up to a million tokens. In the Beijing region the rates are $1.413 and $4.24. Context caching is charged separately: writing to a five-minute cache costs $1.875 per million tokens and to a one-hour cache $3, and reading cached tokens costs $0.075, which makes long, repeated system prompts and documents far cheaper on the second call. The preview's Token Plan route closes on 25 November.
Context, inputs and reasoning
The context window is 1,000,000 tokens, the tier Alibaba Cloud prices. Its documentation for the preview gave up to 393,216 output tokens shared between thinking and the answer, with thinking on by default; Alibaba has not yet published a separate page for the general model, so a team should confirm the output limit in the console. No weights have been published, so it runs only through Alibaba Cloud and the providers that resell it.
Where it sits for a buyer
At $4.50 per million output tokens it is priced below the $7.50 Google charges for Gemini 4 Hydrogen, released four days earlier, though output-heavy reasoning work will spend more tokens than the headline suggests. A team already testing the preview can move its evaluation to the general model before the preview is retired, and should compare the two on its own prompts rather than assume they behave the same.
Key facts
Maker: Alibaba Cloud (Qwen) Released: 16 November 2026 Model code: qwen4-max Context window: 1,000,000 tokens Input: text Price per million tokens (international): $1.50 input, $4.50 output Cache: write $1.875 (5 minutes) or $3 (1 hour), read $0.075
Benchmarks & AI stats
| Benchmark | Official | Community avg |
|---|---|---|
| No benchmark scores yet. Be the first to add one. | ||
“Official” values are editor-approved and feed the ranking. “Community avg” is the mean of member submissions (shown for transparency; it never affects the ranking until an editor approves a value).
Sign in to add a benchmark score for this model.
Qwen4-Max launches
Trust Score
An earned signal from verification, reviews, awards, transparency and engagement - the vendor can't buy it.
Updated 11/17/2026
Request a demo or quote from Qwen4-Max
Reviews
Write the first review of Qwen4-Max
Used it? Your experience helps other buyers decide.
Write a reviewQuestions & answers
No questions yet. Be the first to ask about Qwen4-Max.
Qwen4-Max alternatives
Similar listings buyers compare against this one.
Microsoft AI's real-time speech-to-text model for voice agents, transcribing 60 languages with continuous language detection
Aleph Alpha's open-weight English-German mixture-of-experts model for self-hosting, released under Apache 2.0 in October 2026
Open-weight translation model from Bilibili's Index team, covering 150 languages and following glossary and style instructions
Open-source 2B decision model from AWS's Strands Agents project that picks options and returns scores with a reliability figure
Microsoft AI's high-volume, low-latency variant of MAI-Voice-2.1 for speech in 23 languages, aimed at voice agents
Cantina Security's first open-weight security model, a post-trained GLM-5.3-Flash for vulnerability research and exploit checks
Black Forest Labs' image generation and editing model with bounding-box layouts, multiple reference images and 4K output
More on TrustList
Everything here links back to the same verified catalogue. Pick your next stop.
- Write a reviewHelp other buyers decide
- CompaniesAgencies, consultancies and IT service providers, ranked by verified reviews.
- ProductsSoftware and SaaS with pricing, features, integrations and alternatives.
- ArticlesGuides, comparisons and research from the editorial desk and the community.
- AwardsAnnual recognition decided by verified reviews and an independent jury.
- LaunchesNew products and releases, voted up by the community every day.
- RequestsBuyers describe what they need; vendors respond directly.
- PeopleReviewers, authors and makers with public profiles.
- ComparePut up to four listings side by side before you shortlist.