Alibaba Cloud / Qwen Team model product
Qwen4 Max supports text generation, image and video understanding, and reasoning.
Updated Nov 22, 2026. Default version: Qwen4-Max
This profile uses the model's current scored version. Arena ratings and prices are shown separately.
Benchmark scores for Qwen4-Max.
Benchmark | Score | Rank | Participants | Percentile | Evidence | Evaluated |
|---|
| BenchmarkAPEX-Agents | Score72.10% | Rank01 | Participants16 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkJob Bench | Score68.50% | Rank01 | Participants10 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkWebDev Arena | Score1,750.00 points | Rank02 | Participants3 | Percentile50.00% | EvidenceC | Evaluated |
| BenchmarkBioMysteryBench | Score85.00% | Rank04 | Participants9 | Percentile62.50% | EvidenceC | Evaluated |
| BenchmarkLM Arena Webdev | Score1,750.90 rating | Rank06 | Participants112 | Percentile95.50% | EvidenceA | Evaluated |
| BenchmarkDeepSWE 1.1 | Score75.70% | Rank07 | Participants59 | Percentile89.66% | EvidenceC | Evaluated |
| BenchmarkChartography | Score54.20% | Rank09 | Participants9 | Percentile0.00% | EvidenceC | Evaluated |
| BenchmarkLM Arena Agent Task Outcome Explicit | Score8.82% | Rank09 | Participants50 | Percentile83.67% | EvidenceA | Evaluated |
| BenchmarkLM Arena Agent Leaderboard | Score7.27% | Rank10 | Participants50 | Percentile81.63% | EvidenceA | Evaluated |
| BenchmarkLM Arena Agent Praise Complaint | Score12.99% | Rank10 | Participants50 | Percentile81.63% | EvidenceA | Evaluated |
| BenchmarkLM Arena Agent Steerability | Score8.03% | Rank12 | Participants50 | Percentile77.55% | EvidenceA | Evaluated |
| BenchmarkLM Arena Agent Tool Hallucination | Score1.18% | Rank13 | Participants50 | Percentile75.51% | EvidenceA | Evaluated |
| BenchmarkProgram Bench | Score31.50% | Rank14 | Participants22 | Percentile38.10% | EvidenceC | Evaluated |
| BenchmarkTerminal-Bench 4.0 | Score48.50% | Rank14 | Participants41 | Percentile67.50% | EvidenceC | Evaluated |
| BenchmarkAgents' Last Exam | Score46.70% | Rank16 | Participants33 | Percentile53.13% | EvidenceC | Evaluated |
| BenchmarkHealthBench Professional | Score60.00% | Rank17 | Participants25 | Percentile33.33% | EvidenceC | Evaluated |
| BenchmarkLM Arena Agent Bash Reliability | Score3.35% | Rank18 | Participants50 | Percentile65.31% | EvidenceA | Evaluated |
| BenchmarkHumanity's Last Exam | Score51.70% | Rank22 | Participants119 | Percentile82.20% | EvidenceC | Evaluated |
Preference and agent-evaluation results for the default version.
Arena | Category | Rank | Rating / score | Votes | Observations | Result date |
|---|
| Arenawebdev | Categoryimage to webdev | Rank04 | Rating / score1,674.86 | Votes1,488 | ObservationsN/A | Result date |
| Arenawebdev | Categoryoverall | Rank06 | Rating / score1,750.90 | Votes1,797 | ObservationsN/A | Result date |
| Arenawebdev | Categorywebdev | Rank06 | Rating / score1,750.90 | Votes1,797 | ObservationsN/A | Result date |
| Arenawebdev | Categorywebdev-html | Rank06 | Rating / score1,751.95 | Votes329 | ObservationsN/A | Result date |
| Arenawebdev | Categorywebdev-react | Rank06 | Rating / score1,784.62 | Votes1,028 | ObservationsN/A | Result date |
| Arenaagent task outcome explicit | Categoryoverall | Rank09 | Rating / score0.09 | VotesN/A | Observations6K | Result date |
| Arenaagent | Categoryoverall | Rank10 | Rating / score0.07 | VotesN/A | Observations1.7M | Result date |
| Arenaagent praise complaint | Categoryoverall | Rank10 | Rating / score0.13 | VotesN/A | Observations2.8K | Result date |
| Arenaagent steerability | Categoryoverall | Rank12 | Rating / score0.08 | VotesN/A | Observations51.1K | Result date |
| Arenaagent tool hallucination | Categoryoverall | Rank13 | Rating / score0.01 | VotesN/A | Observations1.5M | Result date |
| Arenaagent bash reliability | Categoryoverall | Rank18 | Rating / score0.03 | VotesN/A | Observations185.4K | Result date |
Official vendor API pricing appears first, followed by individual provider offers.
Provider | Provider model ID | Region | Input / 1M | Output / 1M | Context | Updated |
|---|
| ProviderKenari | Provider model IDqwen4-max:free | Regionglobal | Input / 1MN/A | Output / 1MN/A | Context1M | Updated |
| ProviderNanoGPT | Provider model IDqwen/qwen4-max | Regionglobal | Input / 1M$1.41 | Output / 1M$4.24 | Context1M | Updated |
| ProviderNovita AI | Provider model IDqwen/qwen4-max | Regionglobal | Input / 1M$1.41 | Output / 1M$4.24 | Context1M | Updated |
| ProviderDeep Infra | Provider model IDQwen/Qwen4-Max | Regionglobal | Input / 1M$1.41 | Output / 1M$4.24 | Context1M | Updated |
| ProviderOpenRouter | Provider model IDqwen/qwen4-max | Regionglobal | Input / 1M$1.43 | Output / 1M$4.28 | Context1M | Updated |
| ProviderDevPass (LLM Gateway) | Provider model IDqwen4-max | Regionglobal | Input / 1M$1.45 | Output / 1M$4.25 | Context1M | Updated |
| ProviderVercel AI Gateway | Provider model IDalibaba/qwen4-max | Regionglobal | Input / 1M$1.5 | Output / 1M$4.5 | Context1M | Updated |
| ProviderEmpirioLabs AI | Provider model IDqwen4-max | Regionglobal | Input / 1M$1.5 | Output / 1M$4.5 | Context983.6K | Updated |
| ProviderVenice AI | Provider model IDqwen-4-max | Regionglobal | Input / 1M$1.5 | Output / 1M$4.5 | Context1M | Updated |
| ProviderOpenCode Go | Provider model IDqwen4-max | Regionglobal | Input / 1M$1.5 | Output / 1M$4.5 | Context1M | Updated |
| ProviderEden AI | Provider model IDqwen/qwen4-max | Regionglobal | Input / 1M$1.5 | Output / 1M$4.5 | Context1M | Updated |
| ProviderOfox | Provider model IDqwen/qwen4-max | Regionglobal | Input / 1M$1.5 | Output / 1M$4.5 | Context1M | Updated |
| ProviderLLM Gateway | Provider model IDalibaba/qwen4-max | Regionglobal | Input / 1M$1.5 | Output / 1M$4.5 | Context983.6K | Updated |
| ProviderCharm Hyper | Provider model IDqwen4-max | Regionglobal | Input / 1M$1.5 | Output / 1M$4.5 | Context1M | Updated |
| ProviderCrossModel | Provider model IDqwen/qwen4-max | Regionglobal | Input / 1M$1.6 | Output / 1M$4.8 | Context1M | Updated |
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
Provider-specific output speed and catalog latency for Qwen4-Max. Runtime does not affect the capability score.
No provider-specific speed or latency record is linked to the default version yet.
Output Speed is generated output tokens received per second. Catalog latency is reported separately from observed provider TTFT.
Technical details for the model's default version.
Available versions of this model. The score column identifies the version used in the overall ranking.
Version | Released | LLMBoard | Parameters | Context | Max output | Open weights | License |
|---|
| VersionQwen4-Max | Released | LLMBoard79.01 | Parameters2.4T | Context983.6K | Max output262.1K | Open weightsNo | LicenseProprietary |
| VersionQwen4-Max-Preview | Released | LLMBoardN/A | ParametersN/A | ContextN/A | Max outputN/A | Open weightsNo | LicenseProprietary |
Recommendations prioritize the same model type and family, then the closest LLMBoard score.
Key information about Qwen4 Max and its available data.
Qwen4-Max-Preview is a multimodal large language model in the Qwen family. It supports text generation, image and video understanding, and reasoning.
Data as of 2026-11-21.
Common questions about Qwen4 Max.
Qwen4 Max's default version was released on Nov 16, 2026.
No official standard PAYG price is currently available for Qwen4 Max. The lowest tracked third-party offer starts at $1.41 input and $4.24 output via NanoGPT.
Qwen4 Max was created by Alibaba Cloud / Qwen Team.
The default version has a 983.6K token context window.
No. The default version is not marked as having publicly available weights.
15 provider offerings are linked to the default version.