RULER — Context-length-specific RULER variant, distinct from all other lengths.
| # | Model | Vendor | Best score | Runs | Last seen |
|---|---|---|---|---|---|
| 1 | Nemotron 3 Super 120B A12B | NVIDIA | 95.7 | 1 | 2026-07-07 |
| 2 | NVIDIA-Nemotron-Labs-3-Puzzle-75B-A9B-BF16 | NVIDIA | 94.2 | 1 | 2026-07-07 |
| 3 | MiniMax Text 01 | MiniMax | 92.8 | 1 | 2026-06-06 |
| 4 | Gemini 1.5 Pro | 86.1 | 1 | 2026-06-06 | |
| 5 | Nemotron 3 Ultra 550B A55B Base | NVIDIA | 84.5 | 1 | 2026-06-05 |
| 6 | NVIDIA-Nemotron-3-Ultra-550B-A55B-Base-BF16 | NVIDIA | 84.5 | 1 | 2026-08-24 |
| 7 | Gemini 2.0 Flash Exp | 70.9 | 1 | 2026-06-06 |
Best tracked score per model (default configuration; source-attributed and verification-tiered). Open a model for its full benchmark surface, provenance and pricing.