- MMBench-V1.1-EN 85.8
- MMBench-V1.1-CN 84.7
- MMStar 72.9
- BLINK 65.1
- MUIRBENCH 74.7
- MMMU 68
- MMMU-Pro 57.1
- VideoMMMU 61
- AI2D 87.9
- MathVista 80.7
- WeMath 63.8
- ChartQAPro 59.5
- ChartMuseum 48.8
- OCRBench 84.2
- MMLongBench-Doc 42.4
- RefCOCO-avg (val) 87.4
- OSWorld 14.9
- AndroidWorld 41.7
- WebVoyageSom 69
- Mean of Webquest-SingleQA 72.1 / Webquest-MultiQA 54.7 63.4
- Design2Code 64.7
- Flame-VLM-Code 72.5
- VideoMME (w/) 73.6
- MMVU 59.4
- LVBench 45.1
- MotionBench 59
GLM-4.1V-9B-Thinking
GLM-4.1V-9B-Thinking.
42
Benchmarks with a score
0
Ranked first, out of five or more models
0
In the top three
Z.ai Jul 2025
GLM-4.1V-9B-Thinking
Open weights Small
No Epoch Capabilities Index score yet
- MMMU-Pro 57.1% #59 /72
- ERQA 45.8% #26 /26
- MMVU 59.4% #19 /20
Where it ranks
Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.
Knowledge and language
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| HallusionBench | 63.2% | #15 of 23 | Model card | |
| MMBench-CN v1.1 | 84.7% | #4 of 5 | Model card |
Coding
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Flame-React-Eval | 72.5% | #3 of 3 | Model card |
Agents and tools
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| OSWorld | 14.9% | #11 of 12 | Model card | |
| AndroidWorld | 41.7% | #17 of 18 | Model card | |
| SnakeBench | 13 | #173 of 183 | Official leaderboard | |
| WebVoyager | 69% | #11 of 11 | Model card |
Vision, audio and robotics
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| MMMU | 68% | #41 of 136 | Official leaderboard | |
| MathVista | 80.7% | #24 of 64 | Model card | |
| CVBench | 85% | #5 of 11 | Model card | |
| BLINK | 65.1% | #13 of 28 | Model card | |
| Math-VR | 29% | #14 of 26 | Official leaderboard | |
| MuirBench | 74.7% | #7 of 12 | Model card | |
| AI2D | 87.9% | #26 of 44 | Model card | |
| MMStar | 72.9% | #19 of 30 | Model card | |
| ZeroBench | 19.2% | #17 of 25 | Model card | |
| MMBench-EN v1.1 | 85.8% | #12 of 17 | Model card | |
| Video-MME | 73.6% | #31 of 42 | Model card | |
| MathVerse | 68.4% | #9 of 12 | Model card | |
| We-Math | 63.8% | #9 of 11 | Model card | |
| MMMU-Pro | 57.1% | #59 of 72 | Model card | |
| OCRBench | 84.2% | #23 of 28 | Model card | |
| MATH-Vision (MathVision) | 54.4% | #35 of 42 | Model card | |
| LogicVista | 60.4% | #10 of 12 | Model card | |
| MMLongBench-Doc | 42.4% | #11 of 13 | Model card | |
| RefCOCO / RefCOCO+ / RefCOCOg | 87.4% | #17 of 19 | Model card | |
| VideoMME v2 (with subtitles) | 14% | #29 of 31 | Official leaderboard | |
| DynaMath | 42.5% | #16 of 17 | Model card | |
| MMVU | 59.4% | #19 of 20 | Model card | |
| Video-MMMU | 61% | #29 of 30 | Model card | |
| ERQA | 45.8% | #26 of 26 | Model card | |
| LVBench | 45.1% | #25 of 25 | Model card | |
| MVBench | 68.4% | #15 of 15 | Model card | |
| MotionBench | 59% | #9 of 9 | Model card | |
| ChartQA Pro | 59.5% | #8 of 8 | Model card | |
| TreeBench | 37.5% | #7 of 7 | Model card | |
| All-Angles Bench | 52.7% | #6 of 6 | Model card | |
| ChartMuseum | 48.8% | #5 of 5 | Model card | |
| OmniSpatial | 47.7% | #5 of 5 | Model card | |
| Ref-L4 | 86.8% | #3 of 3 | Model card | |
| WebQuest | 63.4% | #3 of 3 | Model card | |
| GeoBench | 76% | #2 of 2 | Model card |
Its launch post
The benchmarks the lab chose to headline, with the scores it gave.
Its system card
Every result in the longer documents the lab published with it. 14 of them aren't in the launch post.
Capabilities
- MMBench-EN v1.1 85.8
- MMBench-CN v1.1 84.7
- MMStar 72.9
- BLINK 65.1
- MuirBench 74.7
- HallusionBench 63.2
- ZeroBench 19.2
- GeoBench 76
- MMMU 68
- MMMU-Pro 57.1
- MathVista 80.7
- MATH-Vision (MathVision) 54.4
- MathVerse 68.4
- DynaMath 42.5
- LogicVista 60.4
- AI2D 87.9
- We-Math 63.8
- MMLongBench-Doc 42.4
- OCRBench 84.2
- ChartQA Pro 59.5
- ChartMuseum 48.8
- RefCOCO / RefCOCO+ / RefCOCOg 85.3
- TreeBench 37.5
- Ref-L4 86.8
- OmniSpatial 47.7
- CVBench 85
- ERQA 45.8
- All-Angles Bench 52.7
- OSWorld 14.9
- AndroidWorld 41.7
- WebVoyager 69
- WebQuest
- Design2Code 64.7
- Flame-React-Eval 72.5
- Video-MME 73.6
- MMVU 59.4
- Video-MMMU 61
- LVBench 44
- MotionBench 59
- MVBench 68.4