Benchmark
ZEROBench
What is ZEROBench?
ZEROBench is a challenging vision benchmark designed to test models on zero-shot visual understanding tasks. Benchmark Radar tracks 9 reported scores for…
- Released
- 2025-02-13
- Modality
- image
- Evaluates
- multimodal, reasoning, vision
- Openness
- unknown
- Importer
- LLM Stats
- Reported scores
- 9
Source provenance
- Original evidence https://api.zeroeval.com/leaderboard/benchmarks/zerobench?top_n=500
ZEROBench results and reported scores
LLM Stats
| Model | Organization | Reported value | Reported | Evidence |
|---|---|---|---|---|
| Kimi K2.5 | Moonshot AI | 0.11 | 2026-01-27 | evidence |
| Kimi K3 | Moonshot AI | 0.41 | 2026-07-16 | evidence |
| Muse Spark | Meta | 0.33 | 2026-04-08 | evidence |
| Qwen3 VL 235B A22B Thinking | Qwen | 0.04 | 2025-09-22 | evidence |
| Qwen3.5-122B-A10B | Qwen | 0.09 | 2026-02-24 | evidence |
| Qwen3.5-27B | Qwen | 0.1 | 2026-02-24 | evidence |
| Qwen3.5-35B-A3B | Qwen | 0.08 | 2026-02-24 | evidence |
| Seed 2.1 Pro | ByteDance | 0.18 | 2026-06-24 | evidence |
| Seed 2.1 Turbo | ByteDance | 0.11 | 2026-06-24 | 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.