# BU Bench — AI model rankings

Can the AI drive a real web browser to finish tasks — clicking, filling forms, and navigating sites the way a person would? Run by Browser Use on their BU Bench task set. Higher is better.

7 tracked models have a published BU Bench score. Higher is better. Scores come from published lab reports and benchmark sources; recorded sources appear beside the scores.

## Ranking

| Rank | Model | Developer | Score | Source | Released |
| --- | --- | --- | --- | --- | --- |
| 1 | Claude Opus 4.8 | Anthropic | 74% | Lab | May 28 2026 |
| 2 | Gemini 3.6 Flash | Google | 68% | Lab | Jul 21 2026 |
| 3 | GPT-5.6 Sol | OpenAI | 67% | Lab | Jun 26 2026 |
| 4 | Claude Sonnet 4.6 | Anthropic | 62% | Lab | Feb 17 2026 |
| 5 | Gemini 3.5 Flash | Google | 58% | Lab | May 19 2026 |
| 6 | Gemini 3.5 Flash-Lite | Google | 49% | Lab | Jul 21 2026 |
| 7 | GPT-5.4 mini | OpenAI | 36% | Lab | Mar 17 2026 |


---

Canonical page: https://aireleasetracker.com/benchmark/bu-bench
Site index: https://aireleasetracker.com/llms.txt
Source: AI Release Tracker (https://aireleasetracker.com). Most benchmark scores come from lab launch material; gathered results identify the leaderboard that published them.