WebForge-Bench - Tool Usage: leaderboard
WebForge-Bench (2026) tests browser agents on tasks set in automatically generated, reproducible websites with injected real-web noise, controlled for difficulty along seven dimensions. This board is task accuracy in percent on the tool usage domain with screenshot and DOM input; higher is better.
Metric: Accuracy (%) on the tool usage tasks of the 934 WebForge-Bench tasks on generated, self-contained websites with injected real-web noise, up to 50 browser actions in a Chromium GUI, final-state answer checking; multimodal setting: screenshot plus DOM; higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 10 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Pro | 76.6 | #145 |
| 2 | Claude Sonnet 4.5 | 67.4 | #261 |
| 3 | Gemini 3 Flash | 66 | #175 |
| 4 | GPT-5.2 | 63.1 | #185 |
| 5 | Kimi K2.5 | 61.7 | #266 |
| 6 | GPT-5 Mini | 58.2 | #364 |
| 7 | Qwen 3 VL 235B A22B | 48.2 | #485 |
| 8 | Gemini 2.5 Flash Lite | 33.3 | #964 |
| 9 | GPT-5 Nano | 29.8 | #906 |
| 10 | qwen3-omni-30B | 12.1 | #1608 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=webforge-bench-tool-usage · How It Works · Data refreshed daily, snapshot 2026-11-21.