ZEROBench

ZEROBench is a challenging vision benchmark designed to test models on zero-shot visual understanding tasks.

6rows
scoreprimary metric
2026-05-06sampled

Metadata

Metrics

Score, Normalized Score

Latest Results

Rank Subject Score Model Match Provenance Sampled
1 Muse Spark 0.33 Self-reported 2026-05-06
2 Kimi K2.5 0.11 KIMI MoonshotAI: Kimi K2.5
moonshotai-kimi-k2.5
Self-reported 2026-05-06
3 Qwen3.5-27B 0.10 Qwen3.5-27B
qwen-qwen3.5-27b
Self-reported 2026-05-06
4 Qwen3.5-122B-A10B 0.09 Qwen3.5-122B-A10B
qwen-qwen3.5-122b-a10b
Self-reported 2026-05-06
5 Qwen3.5-35B-A3B 0.08 Qwen3.5-35B-A3B
qwen-qwen3.5-35b-a3b
Self-reported 2026-05-06
6 Qwen3 VL 235B A22B Thinking 0.04 Qwen3 VL 235B A22B Thinking
qwen-qwen3-vl-235b-a22b-thinking
Self-reported 2026-05-06