# Mercury 2.5 Benchmark Scores & Performance

> Mercury 2.5 by Inception has 8 source-displayable benchmark rows but no public overall score. It remains unranked.

Canonical page: https://benchlm.ai/models/mercury-2-5

Last updated: September 8, 2026

## Model Details

| Property | Value |
|----------|-------|
| Creator | Inception |
| Source Type | Proprietary |
| Reasoning Type | Reasoning |
| Context Window | 260K |
| Overall Score | Coming soon |
| Overall Rank | Unranked |

## Family & Coverage

- Family: Mercury 2.5
- Variant: base
- Benchmarks covered: 8 of 427
- Sibling models: [Mercury 2.5 Preview](/models/mercury-2-5-preview)
- Related earlier model: [Mercury 2](/models/mercury-2)
- Coverage note: BenchLM currently has partial benchmark coverage for this model, so the overall score is conservative.

## Agentic Benchmarks

| Benchmark | Score |
|-----------|-------|
| [τ³-bench results](/benchmarks/tau3-bench) | 96.0% |
| [DeepSearchQA](/benchmarks/deepsearchqa) | 34.0% |

## Coding Benchmarks

| Benchmark | Score |
|-----------|-------|
| [SciCode](/benchmarks/scicode) | 38% |

## Reasoning Benchmarks

| Benchmark | Score |
|-----------|-------|
| [AA-LCR](/benchmarks/lcr) | 68.0% |

## Knowledge Benchmarks

| Benchmark | Score |
|-----------|-------|
| [GPQA-D](/benchmarks/gpqa-diamond) | 79.0% |
| [AA-Omniscience Accuracy](/benchmarks/omniscienceaccuracy) | 22.0% |
| [AA-Omniscience Hallucination Rate](/benchmarks/omnisciencehallucinationrate) | 67.0% |

## Instruction Following Benchmarks

| Benchmark | Score |
|-----------|-------|
| [IFBench](/benchmarks/ifbench) | 77% |

## Other Inception Models

- [Mercury 2](/models/mercury-2) - Score: 42.34
- [Mercury 2.5 Preview](/models/mercury-2-5-preview) - Score: not computed