HAL Online Mind2Web: leaderboard
Princeton HAL leaderboard for Online Mind2Web: real-world web browsing tasks on live websites requiring navigation, form-filling, and multi-step interaction.
Metric: Accuracy (%). Source: hal.cs.princeton.edu. Saturation forecast: Around November 2027. 22 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5 (Medium) | 42.33 | #98 (GPT-5) |
| 2 | Claude Sonnet 4 (20250514) | 40 | #213 |
| 3 | O3 (Medium) | 39 | #124 (O3) |
| 4 | Claude 3.7 Sonnet (20250219) | 38.33 | #201 |
| 5 | GPT-4.1 (2025-04-14) | 36.33 | #250 |
| 6 | O4 Mini (High) | 32 | #180 (O4 Mini) |
| 7 | O4 Mini (Low) | 31.67 | #180 (O4 Mini) |
| 8 | Gemini 2.0 Flash (001) | 29 | #360 |
| 9 | DeepSeek R1 | 25.33 | #251 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=hal-online-mind2web · How It Works · Data refreshed daily, snapshot 2026-10-20.