multimodal benchmark
MMBench-V1.1 Leaderboard
MMBench-V1.1 is an improved bilingual benchmark for assessing the multimodal capabilities of vision-language models.
Updated Sep 5, 2026
Models20
Model coverage20
MetricScore
EvidenceB
MMBench-V1.1 Ranking
Higher score ranks better on this benchmark.
Sort by
Rank | Model | Score | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelAC | Score92.80% | Percentile100.00% | Participants20 | EvidenceC | Evaluated |
| Rank02 | ModelAC | Score92.80% | Percentile94.74% | Participants20 | EvidenceC | Evaluated |
| Rank03 | ModelAC | Score92.60% | Percentile89.47% | Participants20 | EvidenceC | Evaluated |
| Rank04 | ModelAC | Score92.30% | Percentile84.21% | Participants20 | EvidenceC | Evaluated |
| Rank05 | ModelAC | Score91.50% | Percentile78.95% | Participants20 | EvidenceC | Evaluated |
| Rank06 | ModelAC | Score90.80% | Percentile73.68% | Participants20 | EvidenceC | Evaluated |
| Rank07 | ModelAC | Score90.60% | Percentile68.42% | Participants20 | EvidenceC | Evaluated |
| Rank08 | ModelAC | Score89.90% | Percentile63.16% | Participants20 | EvidenceC | Evaluated |
| Rank09 | ModelAC | Score88.90% | Percentile57.89% | Participants20 | EvidenceC | Evaluated |
| Rank10 | ModelAC | Score87.50% | Percentile52.63% | Participants20 | EvidenceC | Evaluated |
| Rank11 | ModelAC | Score87.00% | Percentile47.37% | Participants20 | EvidenceC | Evaluated |
| Rank12 | ModelAC | Score86.70% | Percentile42.11% | Participants20 | EvidenceC | Evaluated |
| Rank13 | ModelAC | Score85.10% | Percentile36.84% | Participants20 | EvidenceC | Evaluated |
| Rank14 | ModelAC | Score85.00% | Percentile31.58% | Participants20 | EvidenceC | Evaluated |
| Rank15 | ModelAC | Score81.80% | Percentile26.32% | Participants20 | EvidenceC | Evaluated |
| Rank16 | ModelLA | Score81.00% | Percentile21.05% | Participants20 | EvidenceC | Evaluated |
| Rank17 | ModelDE | Score79.30% | Percentile15.79% | Participants20 | EvidenceC | Evaluated |
| Rank18 | ModelDE | Score79.20% | Percentile10.53% | Participants20 | EvidenceC | Evaluated |
| Rank19 | ModelCO | Score68.70% | Percentile5.26% | Participants20 | EvidenceC | Evaluated |
| Rank20 | ModelDE | Score68.30% | Percentile0.00% | Participants20 | EvidenceC | Evaluated |
MMBench-V1.1 Highlights
The leading models and scores on this benchmark.
MMBench-V1.1 Score Distribution
A closer view of the leading scores on this benchmark.
The Top AI Models for MMBench-V1.1
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
What is MMBench-V1.1?
What MMBench-V1.1 measures and how its scores work.
MMBench-V1.1 is a version 1.1 benchmark that evaluates vision-language models using multiple-choice questions in English and Chinese.
It measures performance across diverse vision-language tasks using the Score metric, reported as a ratio.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- MMBench-V1.1
- Modality
- multimodal
- Primary category
- multimodal
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- overall
LLM Stats. Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about MMBench-V1.1.
Which model scores highest on MMBench-V1.1?
Qwen3.5-122B-A10B is currently ranked first with 92.80%.
What are the top three models on MMBench-V1.1?
The current leaders are Qwen3.5-122B-A10B (92.80%), Qwen3.6-35B-A3B (92.80%), and Qwen3.5-27B (92.60%).
Which MMBench-V1.1 model has the lowest official input price?
Qwen2.5-Omni-7B has the lowest matched official input price at $0.10 input / $0.40 output per 1M tokens.
Which models are fastest among MMBench-V1.1 results?
The fastest matched records are Qwen3 VL 4B Thinking (8.77 tok/s via DeepInfra).
Does the highest score result prove overall model quality?
No. This benchmark measures one defined capability or task. The overall LLMBoard score uses a separate aggregation across eligible benchmark evidence.