Mercury 3: benchmark results
Provider: Inception. Released 2026-10-19. Access: API.
Aggregate Rating 1798 ± 22, rank #67 of 1657 rated models, from 30 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Mercury 3 Release - BrowseComp+ | 55.18 | Score (%) | 100 |
| Mercury 3 Release - DeepSearchQA | 71.33 | Score (%) | 100 |
| Mercury 3 Release - GDPval | 1237.2 | Elo | 100 |
| Mercury 3 Release - GPQA Diamond | 89.23 | Accuracy (%) | 100 |
| Mercury 3 Release - IFBench | 76.84 | Score (%) | 100 |
| Mercury 3 Release - MultiChallenge | 60.35 | Score (%) | 100 |
| Mercury 3 Release - SWE-Bench Atlas | 51.08 | Score (%) | 100 |
| Mercury 3 Release - Tau3 Banking | 30.24 | Score (%) | 100 |
| LLM Stats (Tau3 Banking) | 30.24 | Score (%) | 78.6 |
| LLM Stats Score | 36.08 | LLM Stats Score (conservative rating) | 75.5 |
| Design Arena (UI Components) | 1266 | Elo | 73.9 |
| Design Arena (Game Dev) | 1227 | Elo | 68.3 |
Interactive version: theaggregate.ai/model?slug=mercury-3 · How It Works · Data refreshed daily, snapshot 2026-10-22.