R1 — benchmark results
Provider: Other. Access: Open.
Unified ELO 1696 ± 15, rank #337 of 1839 rated models, from 76 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Enhancing Agent Safety Judgment | 94.53 | Original (Seed) (self-reported) | 100 |
| LexGenius | 64.59 | CoT average (self-reported) | 100 |
| MedHELM | 66.25 | Mean Win Rate (self-reported) | 100 |
| PhysicsFinals | 44.3 | Score (self-reported) | 96.9 |
| MixRea | 9.5 | Implicit Information Ignorance Rate (I3R) (self-reported) | 95 |
| Disentangling Language Roles in Multilingual L | 82.2 | JointSuccess (self-reported) | 94.7 |
| TuRTLe Code Completion (Verilator) | 75.99 | Aggregated Score (self-reported) | 90.7 |
| TuRTLe Spec-to-RTL (Icarus Verilog) | 75.53 | Aggregated Score (self-reported) | 90.7 |
| TuRTLe Spec-to-RTL (Verilator) | 75.78 | Aggregated Score (self-reported) | 90.7 |
| Defects4J | 47.5 | Defects4J Plausible @1 (self-reported) | 90.6 |
| OJBench | 26.02 | Score (self-reported) | 88.9 |
| OpenHuEval | 62.31 | Macro Average (computed) (self-reported) | 88.9 |
Interactive version: theaggregate.ai/model?slug=r1 · How It Works · Data refreshed daily, snapshot 2026-08-05.