The Agent Company
Rank #15 of 51 · 14 models scored · top score 42.9
unknown
Strongest on discrimination, weakest on contamination resistance. Caveat: the public test set leaves contamination on the table.
Weighted composite. Each signal is computed from the source-backed leaderboard below.
Does it still separate models?
How far from ceiling / clustered at the top?
Public vs held-out; training-leak risk.
Apples-to-apples, or mixed / vendor-optimized?
How old is the benchmark?
| # | Model | Score | Evidence | Measured |
|---|---|---|---|---|
| 1 | DeepSeek-V3.2-Exp | 42.9 | unverified | 2025-09-29 |
| 2 | Gemini 2.5 Flash (Sep 2025) | 41.1 | unverified | 2025-09-25 |
| 3 | Claude Sonnet 4 | 33.1 | unverified | 2025-05-22 |
| 4 | Claude 3.7 Sonnet | 30.9 | unverified | 2025-02-24 |
| 5 | Gemini 2.5 Pro (May 2025) | 30.3 | unverified | 2025-05-06 |
| 6 | Claude 3.5 Sonnet (October 2024) | 24 | unverified | 2024-10-22 |
| 7 | Gemini 2.0 Flash (Feb 2025) | 11.4 | unverified | 2024-12-11 |
| 8 | GPT-4o (Nov 2024) | 8.6 | unverified | 2024-05-13 |
| 9 | Llama 3.1-405B | 7.4 | unverified | 2024-07-23 |
| 10 | Llama 3.1-70B | 6.9 | unverified | 2024-07-23 |
| 11 | Qwen2.5-72B | 5.7 | unverified | 2024-09-19 |
| 12 | Gemini 1.5 Pro (Sept 2024) | 3.4 | unverified | 2024-09-24 |
| 13 | Amazon Nova Pro | 1.7 | unverified | 2024-12-03 |
| 14 | Qwen2-72B | 1.1 | unverified | 2024-06-07 |