Skip to content

PerceptionBench (Internal)

Moonshot AI's internal benchmark for atomic visual perception capabilities.

3 models

Higher is better.

  1. 1Qwen3.8 MaxAlibaba63.5
  2. 2Kimi K3Moonshot AI58.5
  3. 3dots3-note PreviewDots Studio53.4
Higher is better. Bar colour: blue proprietary, orange open weights, green other.

About this benchmark

Metric
Internal evaluation score
Tasks
Internal atomic visual-perception tasks
Authors
Moonshot AI (2026)

Scores are as reported by model developers or evaluators and may use different settings (for example reasoning effort or tools). Compare with care.