WebsiteBenchWebsiteBench

OpenTable

Rebuilt by Claude Opus 5.5: 150 of 512 hidden tests passed (29.3%)

reference.local/
Reference for OpenTable, HomeReference
rebuild.local/
Rebuild by Claude Opus 5.5, HomeClaude Opus 5.5
    Every model on OpenTable
    ModelPassed%Screenshot
    Claude Opus 5.515029.3
    Claude Opus 59418.4
    GLM 5.3 Flash9318.2
    GPT 6.1 Sol8216.0
    Kimi K38115.8
    DeepSeek V4.1 Flash7915.4
    Grok 4.6356.8
    Qwen 3.8 Max00.0—