LLM Stats leaderboard
6 scores · 6 models · first recorded Oct 2, 2026
higher is better · scale 0–1 · record PinchBench
Current standings — best score per model × subset
| Model | Value (%) | Subset / variant | Test conditions | Reported by | Reported | Document |
|---|---|---|---|---|---|---|
| Nemotron 3 Ultra (550B A55B) | 90% | – | – | NVIDIA | Jun 4, 2026 | //api.zeroeval.com/leaderboard/benchmarks/pinchbench?top_n=500 |
| Nemotron 3.5 Lightning (30B A3B) | 85.4% | – | – | NVIDIA | Aug 11, 2026 | //api.zeroeval.com/leaderboard/benchmarks/pinchbench?top_n=500 |
| MiMo-V2-Omni | 81.2% | – | – | Xiaomi | Mar 18, 2026 | //api.zeroeval.com/leaderboard/benchmarks/pinchbench?top_n=500 |
| MiMo-V2-Pro | 81% | – | – | Xiaomi | Mar 18, 2026 | //api.zeroeval.com/leaderboard/benchmarks/pinchbench?top_n=500 |
| GLM-5V-Turbo | 80.7% | – | – | Z.ai | Apr 2, 2026 | //api.zeroeval.com/leaderboard/benchmarks/pinchbench?top_n=500 |
| LFM2.5-2.6B | 68.2% | – | – | Liquid AI | Aug 4, 2026 | //api.zeroeval.com/leaderboard/benchmarks/pinchbench?top_n=500 |