Balrog
Rank #17 of 51 · 24 models scored · top score 58.1
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 | Gemini 3 Pro | 58.1 | unverified | 2025-11-18 |
| 2 | Gemini 3.1 Pro | 57 | unverified | 2026-02-19 |
| 3 | Gemini 3 Flash | 48.1 | unverified | 2025-12-17 |
| 4 | Grok 4 | 43.6 | unverified | 2025-07-09 |
| 5 | Claude Opus 4.5 | 43.5 | unverified | 2025-11-24 |
| 6 | Gemini 2.5 Pro (Mar 2025) | 43.3 | unverified | 2025-03-25 |
| 7 | DeepSeek-R1 | 34.9 | unverified | 2025-01-20 |
| 8 | Gemini 2.5 Flash (Jun 2025) | 33.5 | unverified | 2025-06-17 |
| 9 | GPT-5 | 32.8 | unverified | 2025-08-07 |
| 10 | Claude 3.5 Sonnet (October 2024) | 32.6 | unverified | 2024-10-22 |
| 11 | GPT-4o (May 2024) | 32.3 | unverified | 2024-05-13 |
| 12 | Claude Haiku 4.5 | 31.2 | unverified | 2025-10-15 |
| 13 | Grok 3 | 29.5 | unverified | 2025-02-17 |
| 14 | Llama 3.1-70B | 27.9 | unverified | 2024-07-23 |
| 15 | Llama 3.2 90B | 27.3 | unverified | 2024-09-24 |
| 16 | Llama 3.3 70B | 23 | unverified | 2024-12-06 |
| 17 | Gemini 1.5 Pro (Sept 2024) | 21 | unverified | 2024-09-24 |
| 18 | Claude 3.5 Haiku | 19.3 | unverified | 2024-10-22 |
| 19 | Mistral NeMo | 17.6 | unverified | 2024-07-18 |
| 20 | GPT-4o mini | 17.4 | unverified | 2024-07-18 |
| 21 | Qwen2.5-72B | 16.2 | unverified | 2024-09-19 |
| 22 | Llama 3.1-8B | 15.1 | unverified | 2024-07-23 |
| 23 | Gemini 1.5 Flash (Sep 2024) | 14.6 | unverified | 2024-05-10 |
| 24 | Phi-4 | 11.6 | unverified | 2024-12-12 |