WebChoreArena: leaderboard
Reproducible browser-agent benchmark of 532 tedious web tasks extending WebArena with massive-memory, calculation, and long-term-memory chores.
Metric: Task success rate (%, full task set). Source: github.com. Source: webchorearena.github.io. Saturation forecast: Rough model projection: around 2026. 8 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | ColorBrowserAgent + gpt-5 | 47.4 | |
| 2 | BrowserGym + gemini-2.5-pro-preview-03-25 | 44.9 | |
| 3 | AgentOccam + gemini-2.5-pro-preview-03-25 | 37.8 | |
| 4 | WEBDART + gpt-5 | 31.1 | |
| 5 | AgentOccam + claude-3-7-sonnet-20250219 | 23.5 | |
| 6 | BrowserGym + claude-3-7-sonnet-20250219 | 23.1 | |
| 7 | AgentOccam + gpt-4o-2024-05-13 | 6.8 | |
| 8 | BrowserGym + gpt-4o-2024-05-13 | 2.6 |
Interactive version: theaggregate.ai/benchmark?slug=webchorearena · How It Works · Data refreshed daily, snapshot 2026-10-20.