The Agent Company — leaderboard
Benchmarks AI agents on realistic software company tasks — coding, project management, data science, HR, and admin work in a simulated internal company environment.
Metric: Overall Score (%). Source: the-agent-company.com. 19 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | OpenHands-0.28.1-claude37-gpt4o | 54.13 |
| 2 | TTE-MatrixAgent-Deepseek-V3.2 | 52.4 |
| 3 | OpenHands-0.28.1-claude37-deepseekv3 | 52.38 |
| 4 | MUSE-gemini-2.5-flash | 51.78 |
| 5 | OpenHands-Versa-claude-sonnet-4 | 43.19 |
| 6 | OpenHands-Versa-claude-3.7-sonnet | 40.18 |
| 7 | OpenHands-0.28.1-gemini-2.5-pro | 39.28 |
| 8 | OpenHands-0.28.1-sonnet-20250219 | 36.41 |
| 9 | OpenHands-0.14.2-sonnet-20241022 | 34.4 |
| 10 | OpenHands-0.14.2-gemini-2.0-flash | 18.96 |
Interactive version: theaggregate.ai/benchmark?slug=the-agent-company · How the rankings work · Data refreshed daily, snapshot 2026-07-22.