Current model · Alibaba Qwen
Qwen4-Max benchmarks, pricing and specs
Qwen4-Max is an Alibaba Qwen model released on November 16, 2026 that takes text, image and video input and produces text. It costs $1.50 per 1M input tokens and $4.50 per 1M output tokens and accepts up to 1,000,000 tokens of context. It scores 1750 Elo on Arena WebDev, #6 of 46 current models (Arena, measured November 15, 2026).
- Per 1M input tokens
- $1.50Listed today
- Per 1M output tokens
- $4.50Listed today
- Context window
- 1M#33 of 71 current models
- Arena WebDev
- 1750 Elo#6 of 46 current models
Data verified November 17, 2026 · Dataset 2026-11-17
Specs
Key facts
- Provider
- Alibaba Qwen
- Model family
- Qwen4
- Release date
- November 16, 2026
- Status
- Current
- Access
- API only
- API model id
- qwen4-max
- Context window
- 1,000,000 tokens
- Max output
- 131,072 tokens
- Input
- text, image, video
- Output
- text
- Task families
- text, multimodal
- Declared capabilities
- reasoning, coding, agentic coding, image understanding, video understanding, chart understanding
Charts
Against every current text model
Where Qwen4-Max lands among the current models of its task, on every independent benchmark, on price and on context.
Where it stands
On Arena WebDev, Qwen4-Max ranks #6 of 46 current text models. 32 of 60 list a lower price per 1M input tokens. Each bar is the share of current models it beats.
Independent benchmarks
Arena WebDev
#6of 46
1,750 Elo · ahead of 89%
Published rating range 1,734 to 1,765
Price and context
Price per 1M input tokens
#33of 60
$1.50 · cheaper than 42%
Price per 1M output tokens
#32of 60
$4.50 · cheaper than 47%
Context window
#33of 71
1M tokens · longer than 21%
Price
Qwen4-Max pricing
It is the #26 most expensive of 60 current text models priced per 1M input tokens.
Benchmarks
Independent benchmark results
Scores come from independent evaluators that ran the model themselves, pinned to a dated snapshot. Rank is among current models in the same snapshot. Alibaba Qwen's own launch figures are in the next section, because they are not measured the same way.
| Benchmark | Score | Rank | Measured | Source |
|---|---|---|---|---|
| Arena WebDev | 1750 EloPublished rating range 1734 to 1765 | #6 of 46 | November 15, 2026 | Arena · 2026-11-17 |
Provider results
Reported by Alibaba Qwen
Alibaba Qwen's own results for Qwen4-Max, beside the rivals it chose, as its table lists them. Labs pick the benchmarks and run them their own way, so compare within one table only. Internal benchmarks and other evaluators' indexes are left out.
Qwen4-Max launch chart, November 16, 2026Qwen Team · 2026-11-17
| Benchmark | Qwen4-Max | Qwen3.8-Max | Claude Opus 5.5 | GPT-6.1 Sol | GPT-6 Astra | Kimi K3.1 Pro |
|---|---|---|---|---|---|---|
| DeepSWE v1.1 | 75.7% | 69.3% | 74.2% | 75.2% | 74.1% | 75.5% |
| Terminal-Bench 4.0 | 48.5% | 38.9% | 59.6% | 56.1% | 59.1% | 67.2% |
| ProgramBench | 31.5% | 28.0% | 65.0% | 52.5% | 55.0% | 25.0% |
| Agents' Last Exam v1.1 | 46.7% | 27.9% | 43.4% | Not listed | 42.6% | 43.4% |
| APEX-Agents v1.1 | 72.1% | 63.3% | 73.5% | 60.0% | 64.7% | 65.5% |
| HealthBench Professional | 60.0% | 39.6% | 59.9% | 65.1% | 66.0% | 57.2% |
| Humanity's Last Exam | 51.7% | 43.9% | 59.8% | 54.0% | 53.6% | 51.1% |
| BioMysteryBench | 85.0% | 84.1% | 81.9% | 84.4% | 78.9% | 80.6% |
| Chartography | 54.2% | 42.0% | 66.3% | 67.8% | 71.1% | 50.0% |
Alibaba's chart from the Qwen4-Max announcement. Qwen3.8-Max is the 0902 refresh, and Kimi K3.1 is Kimi K3.1 Pro: its DeepSWE and BioMysteryBench figures match Moonshot's own card. The APEX-Agents and Agents' Last Exam scores come from those benchmarks' leaderboards; Alibaba ran Chartography for its own two models with a larger image budget. Left out here: CodeArena and AgentArena (Arena ratings), AA-Briefcase and AutomationBench-AA (Artificial Analysis), HLAB-Vals-1.1 (taken from the Vals Index), and JobBench and Video-Index, which we could not confirm are public.
Fit
Strengths, cautions and best uses
- Strengths
- Alibaba's new flagship and the first Qwen4 model: a mixture-of-experts with 2.4T backbone parameters (71B active per token) and 328B N-gram embedding parameters, with native image and video input and a 1,000,000-token context
- $1.50 input, $4.50 output and $0.075 cached input per 1M tokens in Singapore, below Qwen3.8-Max's $2 and $6; Alibaba puts its average cost per Work360 task at $1.67, against $4.43 for Qwen3.8-Max
- In Alibaba's chart it scores 75.7% on DeepSWE v1.1, against 74.2% for Claude Opus 5.5 and 75.5% for Kimi K3.1 Pro, and 46.7% on Agents' Last Exam v1.1, the highest of the five models with a score there
- Arena WebDev rates it 1750 on its first board (November 15, 2026): level with GPT-6.1 Sol (1742) and GPT-6 Astra (1763) within the published ranges, above Kimi K3.1 Pro (1686) and Qwen3.8-Max (1670)
- Up to 131,072 output tokens, or 262,144 in thinking mode (the default), with web search, tool calling, structured output, and OpenAI- and Anthropic-compatible APIs
- Cautions
- API only: Alibaba has not released the weights, and says only that smaller Qwen4 models are coming to open source
- Artificial Analysis scores it 43 on its Intelligence Index against 59 for Claude Opus 5.5 (November 2026), and Alibaba's own chart has Opus 5.5 ahead on Terminal-Bench 4.0 (59.6% against 48.5%) and ProgramBench (65.0% against 31.5%), with Kimi K3.1 Pro at 67.2% on Terminal-Bench
- The recorded rates are the Singapore (international) price; Beijing and the Global regions charge $1.413 input and $4.24 output per 1M tokens, and there is no batch discount
- Arena Text has not rated it yet, and its WebDev rating rests on 1,814 votes so far
- Best for
- Coding and long-horizon agent work at a lower price than Claude Opus 5.5 or GPT-6.1 Sol
- Office deliverables and research over long documents, images and video in one 1M-token context
From benchmark to production
Building on Qwen4-Max?
Public scores rarely hold on your own data. We test it against the alternatives on your real tasks, then build the agent or pipeline around the model that wins.
AI Model Cost AuditLaunch price $100, regular price $150
Alternatives
Alternatives to Qwen4-Max
The closest current text models by price. The cheapest model that scores as well or better is listed first; failing that, the cheapest one within 3% of the benchmark's scale.
On score and price
Qwen4-Max beside the 4 alternatives listed below, on Arena WebDev and on blended price per 1M tokens.
Arena WebDev
Measured November 15, 2026, higher is better
- Qwen4-Max
- 1,750
- Gemini 3.6 Flash
- 1,538
- Gemini 4 Hydrogen
- 1,704
- GLM-5.3
- 1,622
- DeepSeek-V4-Pro
- 1,581
Blended price per 1M tokens
Listed today, lower is better
- Qwen4-Max
- $2.25
- Gemini 3.6 Flash
- $3
- Gemini 4 Hydrogen
- $3
- GLM-5.3
- $2.15
- DeepSeek-V4-Pro
- $1.98
Gemini 3.6 Flash
- Same price per 1M input tokens, $1.50
- 212 points lower on Arena WebDev, 1538 vs 1750
- From Google
Gemini 4 Hydrogen
- Same price per 1M input tokens, $1.50
- 46 points lower on Arena WebDev, 1704 vs 1750
- From Google
GLM-5.3
- 7% cheaper per 1M input tokens, $1.40 vs $1.50
- 128 points lower on Arena WebDev, 1622 vs 1750
- From Z.ai
DeepSeek-V4-Pro
- 12% cheaper per 1M input tokens, $1.32 vs $1.50
- 169 points lower on Arena WebDev, 1581 vs 1750
- From DeepSeek
Further reading
More on Qwen4-Max
FAQ
Qwen4-Max FAQ
How much does Qwen4-Max cost?
Qwen4-Max costs $1.50 per 1M input tokens and $4.50 per 1M output tokens at the standard tier. It is the #26 most expensive of 60 current text models priced per 1M input tokens. Source: Alibaba Cloud, verified November 17, 2026.
What is the context window of Qwen4-Max?
Qwen4-Max accepts up to 1,000,000 tokens of context. It can return up to 131,072 tokens in one response. That ranks #33 of 71 current text models by context length.
How does Qwen4-Max score on independent benchmarks?
Qwen4-Max scores 1750 Elo on Arena WebDev, #6 of 46 current models (Arena, measured November 15, 2026). These are independent evaluations, not Alibaba Qwen's own launch figures.
What benchmark results has Alibaba Qwen published for Qwen4-Max?
In the Qwen4-Max launch chart (November 16, 2026), Alibaba Qwen reports 75.7% on DeepSWE v1.1, 48.5% on Terminal-Bench 4.0, 31.5% on ProgramBench and 46.7% on Agents' Last Exam v1.1. These are Alibaba Qwen's own figures, beside the rivals it chose, not independent measurements.
Is Qwen4-Max open source or API only?
Qwen4-Max is available through Alibaba Qwen's API only; its weights are not published.
What are the best alternatives to Qwen4-Max?
The closest current alternatives by price and measured performance are Gemini 3.6 Flash (same price per 1M input tokens, $1.50), Gemini 4 Hydrogen (same price per 1M input tokens, $1.50), GLM-5.3 (7% cheaper per 1M input tokens, $1.40 vs $1.50) and DeepSeek-V4-Pro (12% cheaper per 1M input tokens, $1.32 vs $1.50).
AI Model Brief
The AI models worth knowing, once a week
What shipped, how it scores on independent benchmarks, what it costs per million tokens and what was deprecated, from the data behind our AI model benchmarks.
We use your address only to send the brief. See our privacy policy.
See this week’s briefSources
Sources
- Qwen4-Max model page (Alibaba Cloud Model Studio) · Alibaba Cloud · official · verified November 17, 2026
- Qwen4-Max: Your New Teammate Has Arrived · Qwen Team · official · verified November 17, 2026
- Alibaba Cloud Model Studio newly released models · Alibaba Cloud · official · verified November 17, 2026
- Arena leaderboard dataset (lmarena-ai/leaderboard-dataset) · Arena · independent · verified November 17, 2026 · CC-BY-4.0
All 291 sources behind the benchmarks, with their versions, licenses and notes, are in the source catalog.
Cite this page
Van Data Team. "Qwen4-Max Benchmarks, Pricing and Specs." AI Model Benchmark Dataset 2026-11-17. https://vandatateam.com/ai-model-benchmarks/qwen4-max
Reuse the data: JSON · CSV · or explore every model in the AI model benchmark comparison.
Production evaluation
Deciding whether Qwen4-Max fits your workload?
Public benchmarks narrow the shortlist. We test the finalists on your own tasks, data and cost limits, then build the routing and review controls to run the winner in production.
- Evaluation cases and acceptance thresholds tied to real work
- Model routing, fallbacks and cost controls
