Multimodal · Benchmark profile
A diagram understanding benchmark focused on scientific and educational visual question answering.
Data verified 21 Jul 2026 · Methodology 1.6.0
Benchmark score on AI2D test split
Qwen3.6-35B-A3B leads at 92.7%, followed by Nemotron 3 Nano Omni 30B A3B (88.5%).
Visual analysis
Switch between model placement, score distribution and descriptive provider averages. Every view uses the same sourced leaderboard.
2 ranked models · higher is better
| Model | Rank | Provider | Score |
|---|---|---|---|
| Qwen3.6-35B-A3B | #1 | Alibaba Cloud | 92.7% |
| Nemotron 3 Nano Omni 30B A3B | #2 | NVIDIA | 88.5% |
One best score per model · higher is better
| Rank | Model | Provider | License | Evidence use | Score |
|---|---|---|---|---|---|
| #1 | Qwen3.6-35B-A3B qwen3-6-35b-a3b | Alibaba Cloud | open | Estimated reference | 92.7% |
| #2 | Nemotron 3 Nano Omni 30B A3B nemotron-3-nano-omni-30b-a3b | NVIDIA | open | Estimated reference | 88.5% |
About AI2D test split
A diagram understanding benchmark focused on scientific and educational visual question answering. Results stay tied to the exact model variant and evaluation system. Multiple systems for the same model use the best published score on this page; overall Lumina scoring uses the median of ranking-eligible rows.
Open benchmark source ↗FAQ
A diagram understanding benchmark focused on scientific and educational visual question answering.
Qwen3.6-35B-A3B by Alibaba Cloud currently leads with 92.7%.
2 models in the LuminaBench cohort have a qualifying score on this benchmark.
Related
