Mercury 2.5: benchmark results
Inception's Mercury 2.5 diffusion LLM: about 1,100 tokens/s with tunable reasoning and parallel tool calls. Served only through Inception's API, Baseten and OpenRouter; no published weights. Provider: Inception. Released 2026-09-08. Access: API.
Aggregate Rating 1610 ± 25, rank #404 of 1640 rated models, from 87 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA Long Context Reasoning | 71.67 | Accuracy (%) | 65.1 |
| Vals AI LegalBench | 83.06 | Accuracy (%) | 60.4 |
| AA-Omniscience Index - Software Engineering (SWE) - Rust | 26 | Omniscience Index | 59.2 |
| AA Humanity's Last Exam | 11.77 | Accuracy (%) | 55.2 |
| LM Market Cap LMC Score | 60.9 | LMC Score (0-100) | 54.8 |
| AA-Omniscience Index - Science, Engineering & Mathematics | -21.1 | Omniscience Index | 51.6 |
| AA-Omniscience Accuracy | 22.65 | Accuracy (%) | 51.4 |
| AA-Omniscience Hallucination Rate | 80.33 | Hallucination Rate (%) | 50.2 |
| Artificial Analysis Intelligence Index | 12.34 | Intelligence Index | 49.6 |
| AA-Omniscience Index - Software Engineering (SWE) - C | -12 | Omniscience Index | 48 |
| AA-Omniscience Index - Software Engineering (SWE) - HTML | -22 | Omniscience Index | 47.6 |
| AA-Omniscience Index - Software Engineering (SWE) - Go | -40 | Omniscience Index | 45.5 |
Interactive version: theaggregate.ai/model?slug=mercury-2-5 · How It Works · Data refreshed daily, snapshot 2026-10-13.