Benchmark
Hypersim Benchmark
What is Hypersim?
Hypersim evaluates 3D grounding and depth understanding in synthetic indoor scenes. Benchmark Radar tracks 4 reported scores for Hypersim, cited by 1 source…
- Modality
- image
- Evaluates
- spatial reasoning, 3d, vision
- Openness
- unknown
- Importer
- LLM Stats
- Reported scores
- 4
Source provenance
- Original evidence https://api.zeroeval.com/leaderboard/benchmarks/hypersim?top_n=500
Hypersim results and reported scores
LLM Stats
| Model | Organization | Reported value | Reported | Evidence |
|---|---|---|---|---|
| Qwen3 VL 235B A22B Thinking | Qwen | 0.11 | 2025-09-22 | evidence |
| Qwen3.5-122B-A10B | Qwen | 0.127 | 2026-02-24 | evidence |
| Qwen3.5-27B | Qwen | 0.13 | 2026-02-24 | evidence |
| Qwen3.5-35B-A3B | Qwen | 0.131 | 2026-02-24 | evidence |
Scores are partitioned by the source that reported them and are never merged into a single cross-source ranking, because the sources measure different things and say so.