BrowserGym VisualWebArena: leaderboard
VisualWebArena's image-grounded web tasks as run on ServiceNow's BrowserGym leaderboard, with vision inputs enabled for models that accept them. Scored as task success rate (%); higher is better. Runs use AgentLab's GenericAgent scaffold unless the row names another framework.
Metric: Success rate (%, GenericAgent unless named). Source: huggingface.co. Saturation forecast: Estimated already saturated. 4 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-4o (2024-08-06) | 26.7 | #731 |
| 2 | Claude 3.5 Sonnet (20241022) | 21 | #635 |
| 3 | GPT-4o Mini (2024-07-18) | 16.9 | #1512 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=browsergym-visualwebarena · How It Works · Data refreshed daily, snapshot 2026-11-12.