MM-MT-Bench
Qwen3-VL-235B-A22B 8.50 leads since September 2025 Models hold multi-turn conversations about images such as charts, documents and photos. GPT-4o grades each answer against a reference on a 1 to 10 scale.
Released with Pixtral 12B. Mistral and Qwen VL reports both use it.
- Made by
- Mistral AI
- Released
- October 2024
- Kind
- General visual questions
- Level
- Everyday
- Scored
- Graded by a model
- Answers
- Open-ended
- Tasks
- All public
- Last in a launch post
- September 2025, dropped before it saturated
- Last model added
- January 2026
Qwen3-VL-235B-A22B leads with 8.50 among models released by January 2026
Best score per model as labs report them in their own cards. Labs pick their own settings, so these compare less cleanly than independent runs.
The newest model with a score here came out in January 2026, and nobody has reported a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:
- Claude Fable 5.2
-
Gemini 4 Carbon - GPT-6 Astra
- Claude Opus 5.5
- and 16 more
- 1 Qwen3-VL-235B-A22B 8.50
- 2
STEP3-VL-10B 8.14 - 3 Qwen2.5-VL-72B 7.60
- 4 Qwen3-Omni-30B-A3B 7.40
- 5
Mistral Small 3.1 7.30 - 6 Qwen2.5-Omni-7B 6.00
Source: Labs’ model cards
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Qwen3-VL-235B-A22B | 8.50 | Sep 2025 | Lab model card | |
| 2 | STEP3-VL-10B | 8.14 | Jan 2026 | Lab model card | |
| 3 | Qwen2.5-VL-72B | 7.60 | Feb 2025 | Lab model card | |
| 4 | Qwen3-Omni-30B-A3B | 7.40 | Sep 2025 | Lab model card | |
| 5 | Mistral Small 3.1 | 7.30 | Mar 2025 | Lab model card | |
| 6 | Qwen2.5-Omni-7B | 6.00 | Mar 2025 | Lab model card |
2 models have held number one in labs’ own numbers since its release in October 2024
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Labs’ model cards
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Labs’ model cards
No launch post has listed it since September 2025
Launch posts from every lab that has listed it, one tick per post. Tall colored ticks include it, and short gray ones leave it out.
| Date | Launch post | Score | As the post writes it |
|---|---|---|---|
| Sep 23, 2025 | Qwen3-VL-235B-A22B | 8.5 | MM_MT_Bench |
| Sep 22, 2025 | Qwen3-Omni-30B-A3B | 7.4 | MM-MT-Bench |
| Mar 27, 2025 | Qwen2.5-Omni-7B | 6 | MM-MT-Bench |
| Mar 17, 2025 | Mistral Small 3.1 | 7.3 | MM MT Bench |
5 system cards report it, from 2 labs
Besides launch posts, we read the system cards, model cards and technical reports labs publish with a release. They report far more, including safety and frontier-risk tests.
| Date | Document | Score | As the document writes it | Section |
|---|---|---|---|---|
| Jan 14, 2026 | STEP3-VL-10B technical report | 8.14 | MM-MT-Bench, SeRe (Table 1) | Capabilities |
| Nov 26, 2025 | Qwen3-VL-235B-A22B technical report | 8.5 | MM-MT-Bench (Table 2) | Capabilities |
| Sep 22, 2025 | Qwen3-Omni-30B-A3B technical report | 7.4 | MM-MT-Bench, Instruct (Table 9). Thinking 8.0 (Table 10) | Capabilities |
| Mar 26, 2025 | Qwen2.5-Omni-7B technical report | 6 | MM-MT-Bench (Table 5) | Capabilities |
| Feb 19, 2025 | Qwen2.5-VL-72B technical report | 7.6 | MM-MT-Bench (Table 3) | Capabilities |
More in Vision
- MMMU In use 11.5K college-level multimodal questions across 30 subjects. Multiple-choice and open accuracy. GPT-5.1 85.4%
- MMMU-Pro In use Harder MMMU: college-level multimodal questions with more options and vision-only inputs to remove text shortcuts. GPT-5.6 Sol 84.6%
- CharXiv In use Chart understanding on real arXiv figures. Reasoning questions require synthesizing information from complex charts. Claude Fable 5 93.5%
- MATH-Vision (MathVision) In use 3,040 competition math problems with visual context. Answer accuracy. Qwen 3.8 Max 97.7%
- MMStar In use Models answer 1,500 human-curated image questions that cannot be solved without the image, across 6 core capabilities. Scored by accuracy. Qwen 3.8 Max 85.9%
- ChartQA In use Models answer human-written and generated questions about chart images that need visual and arithmetic reasoning. Scored by relaxed accuracy with a 5 percent numeric tolerance.
Command A Vision 90.9%