Multimodal & Grounded benchmark
AI2D_TEST leaderboard
AI2D test split. Every model the catalog carries a published AI2D_TEST value for, ranked by that value.
CategoryMultimodal & Grounded
MeasureDiagram-grounded QA
TasksDiagram understanding
DifficultyStructured visual reasoning
Published byQwen3.6 launch benchmarks
A diagram understanding benchmark focused on scientific and educational visual question answering.
AI2D_TEST ranking
3 models with a published AI2D_TEST 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 | Diagram-grounded QA |
|---|---|---|---|
| 1 |  Qwen3.6-35B-A3B | Alibaba | 92.7 |
| 2 |  Nemotron 3 Nano Omni 30B A3B | NVIDIA | 88.5 |
| 3 | ZYZAYA1-VL-8B | Zyphra | 87.5 |
Evidence key: Observed
Rows are ordered by the value Qwen3.6 launch benchmarks 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.