WebsiteBenchWebsiteBench

EveryPlate

Rebuilt by Claude Opus 5.5: 416 of 870 hidden tests passed (47.8%)

reference.local/
Reference for EveryPlate, HomeReference
rebuild.local/
Rebuild by Claude Opus 5.5, HomeClaude Opus 5.5
    Every model on EveryPlate
    ModelPassed%Screenshot
    Claude Opus 5.541647.8
    GPT 6.1 Sol28714.2—
    DeepSeek V4.1 Flash23111.5
    GLM 5.3 Flash1557.7
    Claude Opus 5824.1
    Grok 4.6623.1
    Kimi K3412.0
    Qwen 3.8 Max130.6—