Mercury 3 (High): benchmark results
Provider: Inception. Released 2026-10-19. Access: API.
Aggregate Rating 1720 ± 28, rank #306 of 2170 rated models, from 15 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Vals AI Vibe Code Bench | 60.93 | Accuracy (%) | 52.8 |
| Mercor APEX | 50.2 | Best Pass@1 (%) | 44.6 |
| Vals AI Legal Research Bench | 28.85 | Accuracy (%) | 37.1 |
| Vals AI Tax Agent Bench | 18.17 | Accuracy (%) | 32.5 |
| Vals AI Code Migration v1.1 | 39.26 | Hidden-test Pass Rate (%) | 28.6 |
| AI Coding Daily (OpenCode) - React-TS Code Quality | 15.83 | React-TS Code Quality (max 20) points, LLM-judged rubric sco | 23.8 |
| Vals AI Harvey Legal Agent Bench v1.1 | 5 | Task Resolution (%) | 21.2 |
| AI Coding Daily (OpenCode) - Laravel Code Quality | 16.08 | Laravel Code Quality (max 20) points, LLM-judged rubric scor | 19 |
| Vals AI Legal Index | 62.96 | Legal Index (%) | 18.2 |
| Vals AI Finance Agent v2.1 | 54.71 | Dealbreaker-gated Partial Credit (%) | 14.3 |
| Vals AI Terminal-Bench 4.0 | 3.03 | Accuracy (%) | 13.6 |
| Vals AI (Vals Index) | 43.66 | Accuracy (%) | 12.5 |
Interactive version: theaggregate.ai/model?slug=mercury-3-high · How It Works · Data refreshed daily, snapshot 2026-11-05.