Alibaba Cloud / Qwen Team model product
Qwen2.5 VL 32B
5 VL 32B is a vision-language model from the Qwen family for understanding visual content and generating structured outputs.
Updated Sep 8, 2026. Default version: Qwen2.5 VL 32B Instruct
LLMBoard Score17.0Qwen2.5 VL 32B Instruct
Coverage0%28 benchmark families
Context windowN/ATokens
Official input priceN/AOfficial price unavailable
Qwen2.5 VL 32B Capability Profile
This profile uses the model's current scored version. Arena ratings and prices are shown separately.
Qwen2.5 VL 32B Benchmark Results
Benchmark scores for Qwen2.5 VL 32B Instruct.
Sort by
Benchmark | Score | Rank | Participants | Percentile | Evidence | Evaluated |
|---|
| BenchmarkAndroid Control High_EM | Score69.60% | Rank01 | Participants3 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkInfoVQA | Score83.40% | Rank01 | Participants10 | Percentile100.00% | EvidenceC | Evaluated |
| BenchmarkAITZ_EM | Score83.10% | Rank02 | Participants3 | Percentile50.00% | EvidenceC | Evaluated |
| BenchmarkAndroid Control Low_EM | Score93.30% | Rank02 | Participants3 | Percentile50.00% | EvidenceC | Evaluated |
| BenchmarkMMBench-Video | Score1.93% | Rank02 | Participants3 | Percentile50.00% | EvidenceC | Evaluated |
| BenchmarkDocVQA | Score94.80% | Rank06 | Participants28 | Percentile81.48% | EvidenceC | Evaluated |
| BenchmarkAndroidWorld_SR | Score22.00% | Rank08 | Participants8 | Percentile0.00% | EvidenceC | Evaluated |
| BenchmarkMBPP | Score84.00% | Rank08 | Participants33 | Percentile78.13% | EvidenceC | Evaluated |
| BenchmarkOCRBench-V2 (zh) | Score59.10% | Rank08 | Participants11 | Percentile30.00% | EvidenceC | Evaluated |
| BenchmarkVideoMME w sub. | Score77.90% | Rank08 | Participants10 | Percentile22.22% | EvidenceC | Evaluated |
| BenchmarkVideoMME w/o sub. | Score70.50% | Rank09 | Participants10 | Percentile11.11% | EvidenceC | Evaluated |
| BenchmarkCharadesSTA | Score54.20% | Rank11 | Participants12 | Percentile9.09% | EvidenceC | Evaluated |
| BenchmarkHumanEval | Score91.50% | Rank11 | Participants66 | Percentile84.62% | EvidenceC | Evaluated |
| BenchmarkScreenSpot | Score88.50% | Rank11 | Participants16 | Percentile33.33% | EvidenceC | Evaluated |
| BenchmarkOCRBench-V2 (en) | Score57.20% | Rank12 | Participants14 | Percentile15.38% | EvidenceC | Evaluated |
| BenchmarkCC-OCR | Score77.10% | Rank15 | Participants18 | Percentile17.65% | EvidenceC | Evaluated |
| BenchmarkMATH | Score82.20% | Rank18 | Participants71 | Percentile75.71% | EvidenceC | Evaluated |
| BenchmarkMathVista-Mini | Score74.70% | Rank18 | Participants25 | Percentile29.17% | EvidenceC | Evaluated |
| BenchmarkMMStar | Score69.50% | Rank18 | Participants25 | Percentile29.17% | EvidenceC | Evaluated |
| BenchmarkOSWorld | Score5.92% | Rank20 | Participants20 | Percentile0.00% | EvidenceC | Evaluated |
| BenchmarkLVBench | Score49.00% | Rank24 | Participants28 | Percentile14.81% | EvidenceC | Evaluated |
| BenchmarkScreenSpot Pro | Score39.40% | Rank25 | Participants26 | Percentile4.00% | EvidenceC | Evaluated |
| BenchmarkMathVision | Score38.40% | Rank31 | Participants35 | Percentile11.76% | EvidenceC | Evaluated |
| BenchmarkMMMU | Score70.00% | Rank34 | Participants64 | Percentile47.62% | EvidenceC | Evaluated |
| BenchmarkMMMU-Pro | Score49.50% | Rank61 | Participants69 | Percentile11.76% | EvidenceC | Evaluated |
| BenchmarkMMLU | Score78.40% | Rank70 | Participants101 | Percentile31.00% | EvidenceC | Evaluated |
| BenchmarkLM Arena Vision | Score1,152.15 rating | Rank79 | Participants107 | Percentile26.42% | EvidenceA | Evaluated |
| BenchmarkLM Arena Vision Style Control | Score1,119.02 rating | Rank91 | Participants107 | Percentile15.09% | EvidenceA | Evaluated |
| BenchmarkMMLU-Pro | Score68.80% | Rank95 | Participants138 | Percentile31.39% | EvidenceC | Evaluated |
| BenchmarkGPQA | Score46.00% | Rank202 | Participants247 | Percentile18.29% | EvidenceC | Evaluated |
Qwen2.5 VL 32B Arena Results
Preference and agent-evaluation results for the default version.
Sort by
Arena | Category | Rank | Rating / score | Votes | Observations | Result date |
|---|
| Arenavision | Categoryoverall | Rank79 | Rating / score1,152.15 | Votes1,490 | ObservationsN/A | Result date |
| Arenavision | Categoryenglish | Rank80 | Rating / score1,158.30 | Votes870 | ObservationsN/A | Result date |
| Arenavision style control | Categoryenglish | Rank90 | Rating / score1,123.31 | Votes870 | ObservationsN/A | Result date |
| Arenavision style control | Categoryoverall | Rank91 | Rating / score1,119.02 | Votes1,490 | ObservationsN/A | Result date |
Qwen2.5 VL 32B Pricing
Official vendor API pricing appears first, followed by individual provider offers.
- Official API
- N/A
- Official provider
- N/A
- Lowest third-party
- From $0.05 input, $0.22 output per 1M via IO.NET
- Tracked offerings
- 2
Sort by
Provider | Provider model ID | Region | Input / 1M | Output / 1M | Context | Updated |
|---|
| ProviderIO.NET | Provider model IDQwen/Qwen2.5-VL-32B-Instruct | Regionglobal | Input / 1M$0.05 | Output / 1M$0.22 | Context32K | Updated |
| ProviderMeganova | Provider model IDQwen/Qwen2.5-VL-32B-Instruct | Regionglobal | Input / 1M$0.20 | Output / 1M$0.60 | Context16.4K | Updated |
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
Qwen2.5 VL 32B Runtime Performance
Provider-specific output speed and catalog latency for Qwen2.5 VL 32B Instruct. 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.
Qwen2.5 VL 32B Specifications
Technical details for the model's default version.
- Version
- Qwen2.5 VL 32B Instruct
- Released
- Feb 28, 2025
- Knowledge cutoff
- Unknown
- Parameters
- 33.5B
- Context window
- N/A
- Max output
- N/A
- Inputs
- image, text
- Outputs
- text
- Open weights
- Yes
- License
- Apache 2.0
Qwen2.5 VL 32B 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 |
|---|
| VersionQwen2.5 VL 32B Instruct | Released | LLMBoard16.96 | Parameters33.5B | ContextN/A | Max outputN/A | Open weightsYes | LicenseApache 2.0 |
Models similar to Qwen2.5 VL 32B
Recommendations prioritize the same model type and family, then the closest LLMBoard score.
AC
Qwen3 VL 4B
Alibaba Cloud / Qwen Team
AC
Qwen2.5 72B
Alibaba Cloud / Qwen Team
AC
QvQ 72B
Alibaba Cloud / Qwen Team
AC
Qwen2.5 32B
Alibaba Cloud / Qwen Team
AC
Qwen2.5 VL 72B
Alibaba Cloud / Qwen Team
AC
Qwen3 VL 4B Thinking
Alibaba Cloud / Qwen Team
What is Qwen2.5 VL 32B?
Key information about Qwen2.5 VL 32B and its available data.
5 VL 32B is a vision-language model from Alibaba Cloud / Qwen Team. 5-VL supports visual understanding of objects, text, charts, and layouts, along with visual agent capabilities, long-video comprehension, visual localization, and structured output generation.
Data as of 2026-09-08.
FAQ
Common questions about Qwen2.5 VL 32B.
When was Qwen2.5 VL 32B released?
Qwen2.5 VL 32B's default version was released on Feb 28, 2025.
How much does Qwen2.5 VL 32B cost?
No official standard PAYG price is currently available for Qwen2.5 VL 32B. The lowest tracked third-party offer starts at $0.05 input and $0.22 output via IO.NET.
Who created Qwen2.5 VL 32B?
Qwen2.5 VL 32B was created by Alibaba Cloud / Qwen Team.
What is the context window for Qwen2.5 VL 32B?
A context window is not available for the default version.
Is Qwen2.5 VL 32B open weight?
Yes. The default version is marked as open weight under Apache 2.0.
How many API providers offer Qwen2.5 VL 32B?
2 provider offerings are linked to the default version.