Alibaba Cloud / Qwen Team model product
Qwen3 235B A22B Thinking
Qwen3 235B A22B Thinking is a thinking-enabled Mixture-of-Experts (MoE) model with 235B total parameters and 22B activated parameters.
Updated Sep 8, 2026. Default version: Qwen3-235B-A22B-Thinking-2507
LLMBoard Score47.3Qwen3-235B-A22B-Thinking-2507
Coverage80%24 benchmark families
Context window262.1KTokens
Official input priceN/AOfficial price unavailable
Qwen3 235B A22B Thinking Capability Profile
This profile uses the model's current scored version. Arena ratings and prices are shown separately.
Qwen3 235B A22B Thinking Benchmark Results
Benchmark scores for Qwen3-235B-A22B-Thinking-2507.
Sort by
Benchmark | Score | Rank | Participants | Percentile | Evidence | Evaluated |
|---|
| BenchmarkCFEval | Score2,134.00 points | Rank01 | Participants2 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkMulti-IF | Score80.60% | Rank01 | Participants23 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkWritingBench | Score88.30% | Rank01 | Participants15 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkLiveBench 20241125 | Score78.40% | Rank02 | Participants14 | Percentile92.31% | EvidenceC | Evaluated |
| BenchmarkArena-Hard v2 | Score79.70% | Rank03 | Participants19 | Percentile88.89% | EvidenceC | Evaluated |
| BenchmarkCreative Writing v3 | Score86.10% | Rank03 | Participants13 | Percentile83.33% | EvidenceC | Evaluated |
| BenchmarkBFCL-v3 | Score71.90% | Rank06 | Participants19 | Percentile72.22% | EvidenceC | Evaluated |
| BenchmarkOJBench | Score32.50% | Rank06 | Participants9 | Percentile37.50% | EvidenceC | Evaluated |
| BenchmarkMMLU-Redux | Score93.80% | Rank07 | Participants48 | Percentile87.23% | EvidenceC | Evaluated |
| BenchmarkInclude | Score81.00% | Rank08 | Participants31 | Percentile76.67% | EvidenceC | Evaluated |
| BenchmarkMMLU-ProX | Score81.00% | Rank08 | Participants32 | Percentile77.42% | EvidenceC | Evaluated |
| BenchmarkPolyMATH | Score60.10% | Rank08 | Participants23 | Percentile68.18% | EvidenceC | Evaluated |
| BenchmarkSuperGPQA | Score64.90% | Rank11 | Participants34 | Percentile69.70% | EvidenceC | Evaluated |
| BenchmarkHMMT25 | Score83.90% | Rank12 | Participants28 | Percentile59.26% | EvidenceC | Evaluated |
| BenchmarkTau2 Airline | Score58.00% | Rank15 | Participants24 | Percentile39.13% | EvidenceC | Evaluated |
| BenchmarkTau2 Retail | Score71.90% | Rank17 | Participants27 | Percentile38.46% | EvidenceC | Evaluated |
| BenchmarkTAU-bench Airline | Score46.00% | Rank17 | Participants23 | Percentile27.27% | EvidenceC | Evaluated |
| BenchmarkTAU-bench Retail | Score67.80% | Rank17 | Participants25 | Percentile33.33% | EvidenceC | Evaluated |
| BenchmarkMMLU-Pro | Score84.40% | Rank28 | Participants138 | Percentile80.29% | EvidenceC | Evaluated |
| BenchmarkIFEval | Score87.80% | Rank29 | Participants68 | Percentile58.21% | EvidenceC | Evaluated |
| BenchmarkLiveCodeBench v6 | Score74.10% | Rank31 | Participants62 | Percentile50.82% | EvidenceC | Evaluated |
| BenchmarkTau2 Telecom | Score45.60% | Rank32 | Participants36 | Percentile11.43% | EvidenceC | Evaluated |
| BenchmarkAIME 2025 | Score92.30% | Rank38 | Participants119 | Percentile68.64% | EvidenceC | Evaluated |
| BenchmarkHumanity's Last Exam | Score18.20% | Rank70 | Participants103 | Percentile32.35% | EvidenceC | Evaluated |
| BenchmarkGPQA | Score81.10% | Rank88 | Participants247 | Percentile64.63% | EvidenceC | Evaluated |
| BenchmarkLM Arena Text | Score1,414.76 rating | Rank89 | Participants210 | Percentile57.89% | EvidenceA | Evaluated |
| BenchmarkLM Arena Text Style Control | Score1,399.89 rating | Rank109 | Participants210 | Percentile48.33% | EvidenceA | Evaluated |
Qwen3 235B A22B Thinking Arena Results
Preference and agent-evaluation results for the default version.
Sort by
Arena | Category | Rank | Rating / score | Votes | Observations | Result date |
|---|
| Arenatext | Categoryindustry medicine and healthcare | Rank45 | Rating / score1,455.54 | Votes509 | ObservationsN/A | Result date |
| Arenatext | Categorychinese | Rank48 | Rating / score1,482.61 | Votes540 | ObservationsN/A | Result date |
| Arenatext | Categoryexpert | Rank48 | Rating / score1,459.55 | Votes416 | ObservationsN/A | Result date |
| Arenatext | Categoryjapanese | Rank62 | Rating / score1,387.53 | Votes432 | ObservationsN/A | Result date |
| Arenatext | Categoryindustry mathematical | Rank64 | Rating / score1,437.04 | Votes473 | ObservationsN/A | Result date |
| Arenatext style control | Categoryexpert | Rank65 | Rating / score1,465.28 | Votes416 | ObservationsN/A | Result date |
| Arenatext | Categorykorean | Rank68 | Rating / score1,374.92 | Votes204 | ObservationsN/A | Result date |
| Arenatext style control | Categoryjapanese | Rank70 | Rating / score1,378.39 | Votes432 | ObservationsN/A | Result date |
| Arenatext | Categorygerman | Rank79 | Rating / score1,406.40 | Votes210 | ObservationsN/A | Result date |
| Arenatext | Categoryindustry business and management and financial operations | Rank81 | Rating / score1,411.76 | Votes1,526 | ObservationsN/A | Result date |
| Arenatext | Categoryindustry life and physical and social science | Rank81 | Rating / score1,433.21 | Votes1,491 | ObservationsN/A | Result date |
| Arenatext | Categoryenglish | Rank82 | Rating / score1,429.11 | Votes3,891 | ObservationsN/A | Result date |
| Arenatext style control | Categorykorean | Rank82 | Rating / score1,361.58 | Votes204 | ObservationsN/A | Result date |
| Arenatext | Categoryhard prompts english | Rank84 | Rating / score1,431.13 | Votes1,717 | ObservationsN/A | Result date |
| Arenatext style control | Categoryindustry medicine and healthcare | Rank86 | Rating / score1,438.29 | Votes509 | ObservationsN/A | Result date |
| Arenatext | Categorypolish | Rank87 | Rating / score1,389.62 | Votes784 | ObservationsN/A | Result date |
| Arenatext style control | Categorygerman | Rank87 | Rating / score1,398.31 | Votes210 | ObservationsN/A | Result date |
| Arenatext style control | Categorychinese | Rank88 | Rating / score1,455.95 | Votes540 | ObservationsN/A | Result date |
| Arenatext | Categorycreative writing | Rank89 | Rating / score1,383.71 | Votes1,069 | ObservationsN/A | Result date |
| Arenatext | Categoryexclude ties | Rank89 | Rating / score1,401.15 | Votes6,202 | ObservationsN/A | Result date |
| Arenatext | Categoryoverall | Rank89 | Rating / score1,414.76 | Votes8,800 | ObservationsN/A | Result date |
| Arenatext style control | Categoryindustry mathematical | Rank89 | Rating / score1,425.56 | Votes473 | ObservationsN/A | Result date |
| Arenatext | Categorymath | Rank90 | Rating / score1,413.41 | Votes479 | ObservationsN/A | Result date |
| Arenatext | Categoryrussian | Rank90 | Rating / score1,400.96 | Votes532 | ObservationsN/A | Result date |
| Arenatext | Categoryindustry software and it services | Rank91 | Rating / score1,431.10 | Votes2,909 | ObservationsN/A | Result date |
| Arenatext | Categoryhard prompts | Rank92 | Rating / score1,415.21 | Votes3,810 | ObservationsN/A | Result date |
| Arenatext | Categoryindustry writing and literature and language | Rank92 | Rating / score1,387.23 | Votes1,851 | ObservationsN/A | Result date |
| Arenatext | Categorynon english | Rank93 | Rating / score1,397.19 | Votes4,901 | ObservationsN/A | Result date |
| Arenatext style control | Categoryrussian | Rank94 | Rating / score1,406.09 | Votes532 | ObservationsN/A | Result date |
| Arenatext | Categoryindustry legal and government | Rank95 | Rating / score1,415.45 | Votes531 | ObservationsN/A | Result date |
Qwen3 235B A22B Thinking Pricing
Official vendor API pricing appears first, followed by individual provider offers.
- Official API
- N/A
- Official provider
- N/A
- Lowest third-party
- From $0.20 input, $0.60 output per 1M via submodel
- Tracked offerings
- 11
Sort by
Provider | Provider model ID | Region | Input / 1M | Output / 1M | Context | Updated |
|---|
| ProvideriFlow | Provider model IDqwen3-235b-a22b-thinking-2507 | Regionglobal | Input / 1MN/A | Output / 1MN/A | Context256K | Updated |
| ProviderModelScope | Provider model IDQwen/Qwen3-235B-A22B-Thinking-2507 | Regionglobal | Input / 1MN/A | Output / 1MN/A | Context262.1K | Updated |
| Providersubmodel | Provider model IDQwen/Qwen3-235B-A22B-Thinking-2507 | Regionglobal | Input / 1M$0.20 | Output / 1M$0.60 | Context262.1K | Updated |
| ProviderOpenRouter | Provider model IDqwen/qwen3-235b-a22b-thinking-2507 | Regionglobal | Input / 1M$0.23 | Output / 1M$2.3 | Context131.1K | Updated |
| ProviderHugging Face | Provider model IDQwen/Qwen3-235B-A22B-Thinking-2507 | Regionglobal | Input / 1M$0.30 | Output / 1M$3 | Context262.1K | Updated |
| ProviderJiekou.AI | Provider model IDqwen/qwen3-235b-a22b-thinking-2507 | Regionglobal | Input / 1M$0.30 | Output / 1M$3 | Context131.1K | Updated |
| ProviderNovitaAI | Provider model IDqwen/qwen3-235b-a22b-thinking-2507 | Regionglobal | Input / 1M$0.30 | Output / 1M$3 | Context131.1K | Updated |
| ProviderDevPass (LLM Gateway) | Provider model IDqwen3-235b-a22b-thinking-2507 | Regionglobal | Input / 1M$0.30 | Output / 1M$3 | Context262K | Updated |
| ProviderVercel AI Gateway | Provider model IDalibaba/qwen3-235b-a22b-thinking | Regionglobal | Input / 1M$0.40 | Output / 1M$4 | Context131.1K | Updated |
| ProviderVenice AI | Provider model IDqwen3-235b-a22b-thinking-2507 | Regionglobal | Input / 1M$0.45 | Output / 1M$3.5 | Context128K | Updated |
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
Qwen3 235B A22B Thinking Runtime Performance
Provider-specific output speed and catalog latency for Qwen3-235B-A22B-Thinking-2507. Runtime does not affect the capability score.
No runtime data
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.
Qwen3 235B A22B Thinking Specifications
Technical details for the model's default version.
- Version
- Qwen3-235B-A22B-Thinking-2507
- Released
- Jul 25, 2025
- Knowledge cutoff
- Unknown
- Parameters
- 235B
- Context window
- 262.1K
- Max output
- 131.1K
- Inputs
- text
- Outputs
- text
- Open weights
- Yes
- License
- Apache 2.0
Qwen3 235B A22B Thinking Versions
Available versions of this model. The score column identifies the version used in the overall ranking.
Sort by
Version | Released | LLMBoard | Parameters | Context | Max output | Open weights | License |
|---|
| VersionQwen3-235B-A22B-Thinking-2507 | Released | LLMBoard47.32 | Parameters235B | Context262.1K | Max output131.1K | Open weightsYes | LicenseApache 2.0 |
Models similar to Qwen3 235B A22B Thinking
Recommendations prioritize the same model type and family, then the closest LLMBoard score.
AC
Qwen3 VL 235B A22B Thinking
Alibaba Cloud / Qwen Team
AC
Qwen3.5 35B A3B
Alibaba Cloud / Qwen Team
AC
Qwen3 Max
Alibaba Cloud / Qwen Team
AC
Qwen3 VL 235B A22B
Alibaba Cloud / Qwen Team
AC
Qwen3 235B A22B
Alibaba Cloud / Qwen Team
AC
Qwen3.6 35B A3B
Alibaba Cloud / Qwen Team
What is Qwen3 235B A22B Thinking?
Key information about Qwen3 235B A22B Thinking and its available data.
Qwen3 235B A22B Thinking is an LLM with 94 layers, 128 experts, 8 activated experts, and a 262K native context length. It supports only thinking mode with automatic <think> tag inclusion.
Data as of 2026-09-08.
FAQ
Common questions about Qwen3 235B A22B Thinking.
When was Qwen3 235B A22B Thinking released?
Qwen3 235B A22B Thinking's default version was released on Jul 25, 2025.
How much does Qwen3 235B A22B Thinking cost?
No official standard PAYG price is currently available for Qwen3 235B A22B Thinking. The lowest tracked third-party offer starts at $0.20 input and $0.60 output via submodel.
Who created Qwen3 235B A22B Thinking?
Qwen3 235B A22B Thinking was created by Alibaba Cloud / Qwen Team.
What is the context window for Qwen3 235B A22B Thinking?
The default version has a 262.1K token context window.
Is Qwen3 235B A22B Thinking open weight?
Yes. The default version is marked as open weight under Apache 2.0.