PerceptionBenchBrowse 284 families
PerceptionBench
Can the model accurately perceive visual evidence before reasoning or outside knowledge is required?
Version / measure
- Domain
- Vision
- Ranked models
- 5
- Reported values
- 5
- Reports
- 1 · lab
- Direction
- Higher is better
- Status
- Latest stable
Rankings
Best reported score per model across effort settings
| # | Model | Best setting | Score |
|---|---|---|---|
| 1 | GPT-5.6 SolOpenAI | maxcost n/a1 value · Jul 17 · source | 59.7% |
| 2 | Kimi K3Moonshot AIOW | maxcost n/a1 value · Jul 17 · source | 58.5% |
| 3 | Claude Fable 5Anthropic | maxcost n/a1 value · Jul 17 · source | 57.2% |
| 4 | GPT-5.5OpenAI | xhighcost n/a1 value · Jul 17 · source | 55.8% |
| 5 | Claude Opus 4.8Anthropic | maxcost n/a1 value · Jul 17 · source | 47.2% |
Effort curve
Score vs. reported cost. Lines connect complete sweeps only.
No cost-linked effort sweep for this version.
Definition and comparison boundarypublic methodology
Can the model accurately perceive visual evidence before reasoning or outside knowledge is required?
Three thousand verified questions spanning ten atomic visual-perception capabilities. Higher is better. The value is the percentage reported in this lab's table.
Harness, tools, prompts, attempt count, and benchmark version can materially change the result. Compare within one reporting context.
Method / source ↗