Multimodal & Grounded benchmark
V* leaderboard
Every model the catalog carries a published V* value for, ranked by that value.
CategoryMultimodal & Grounded
MeasureVision-centric reasoning benchmark
TasksFrontier multimodal reasoning tasks
DifficultyFrontier multimodal
Published byGLM-5V-Turbo
A vision-centric benchmark for high-level multimodal reasoning and perception quality.
V* ranking
11 models with a published V* value, ordered by that value, highest first. Models the source has not scored on this benchmark are not listed — they are unmeasured, not last.
| Rank | Model | Provider | Vision-centric reasoning benchmark |
|---|---|---|---|
| 1 |  Kimi K2.6 | Moonshot AI | 96.9 |
| 1 |  Qwen3.6 Plus | Alibaba | 96.9 |
| 3 | Qwen3.5 397B | Alibaba | 95.8 |
| 4 |  Step 3.7 Flash | StepFun | 95.3 |
| 5 | Qwen3.6-27B | Alibaba | 94.7 |
| 6 | Qwen3.5-27B | Alibaba | 93.7 |
| 7 | Qwen3.5-122B-A10B | Alibaba | 93.2 |
| 8 | Qwen3.5-35B-A3B | Alibaba | 92.7 |
| 9 |  Gemini 3 Pro | 88.0 | |
| 10 |  GPT-5.2 | OpenAI | 75.9 |
| 11 |  Claude Opus 4.5 | Anthropic | 67.0 |
Evidence key: Observed
Rows are ordered by the value GLM-5V-Turbo published, highest first. The source does not state whether a higher value is the better result, so this page does not either: for a benchmark that measures a rate of failure, read the table from the bottom.