# Mercury 2.5 vs Mercury 3

> Side-by-side benchmark comparison for Mercury 2.5 and Mercury 3 across agentic, coding, multimodal, knowledge, reasoning, multilingual, and math tasks.

- Canonical page: https://benchlm.ai/compare/mercury-2-5-vs-mercury-3
- Last updated: October 19, 2026

- Shared sourced benchmarks: 4
- HTML indexing: indexable
- Ranking lane: BenchAlign v5.8

## Quick Verdict

Mercury 2.5 leads by point estimate. Conditional score ranges do not establish rank confidence. Choose using the category evidence and your own workload tests.

## Summary

- Mercury 2.5 has the higher overall point estimate 33.98 to 0. Conditional score ranges do not establish rank confidence.
- The biggest single benchmark swing is τ³-bench results in Agentic, with scores of 96.0% and 30.2%.
- Mercury 2.5 is the cheaper option on output tokens, which matters if you expect large responses or heavy interactive use.
- Mercury 3 also has the larger context window at 1M.

## Model Snapshot

| Property | Mercury 2.5 | Mercury 3 |
|----------|----------|----------|
| Creator | Inception | Inception |
| Type | Proprietary | Proprietary |
| Reasoning | Reasoning | Reasoning |
| Context | 260K | 1M |
| Overall Score | 33.98 | null |
| Benchmarks Covered | 6 | 8 |
| Pricing (input/output) | $0.04 / $0.15 | $0.10 / $0.35 |

## Category Breakdown

### Agentic

- Winner: Not comparable
- Mercury 2.5 public-lane score: 12.2 (Supported · #121/130)
- Mercury 3 public-lane score: Coming soon

| Benchmark | Mercury 2.5 | Mercury 3 | Winner |
|-----------|-----------|-----------|--------|
| τ³-bench results | 96.0% | 30.2% | Mercury 2.5 |
| DeepSearchQA | 34.0% | 71.3% | Mercury 3 |
| Terminal-Bench 2.1 (Vals) | 34.5% | Coming soon | Coming soon |
| BrowseComp+ | Coming soon | 55.18% | Coming soon |

### Coding

- Winner: Not comparable
- Mercury 2.5 public-lane score: 16.1 (Estimated · #131/151)
- Mercury 3 public-lane score: Coming soon

| Benchmark | Mercury 2.5 | Mercury 3 | Winner |
|-----------|-----------|-----------|--------|
| SciCode | 38% | Coming soon | Coming soon |
| DeepSWE | Coming soon | 46.3% | Coming soon |

### Reasoning

- Winner: Not comparable
- Mercury 2.5 public-lane score: 71.7 (Unranked · 2 rankable rows)
- Mercury 3 public-lane score: Coming soon

| Benchmark | Mercury 2.5 | Mercury 3 | Winner |
|-----------|-----------|-----------|--------|
| MRCRv2 | Coming soon | 75.5% | Coming soon |

### Knowledge

- Winner: Not comparable
- Mercury 2.5 public-lane score: 34.5 (Supported · #138/186)
- Mercury 3 public-lane score: Coming soon

| Benchmark | Mercury 2.5 | Mercury 3 | Winner |
|-----------|-----------|-----------|--------|
| GPQA-D | 79.0% | 89.2% | Mercury 3 |

### Instruction Following

- Winner: Not comparable
- Mercury 2.5 public-lane score: 81.1 (#58/148)
- Mercury 3 public-lane score: Coming soon

| Benchmark | Mercury 2.5 | Mercury 3 | Winner |
|-----------|-----------|-----------|--------|
| IFBench | 77% | 76.8% | Mercury 2.5 |
| MultiChallenge (provider reports) | Coming soon | 60.35% | Coming soon |

## FAQ

### Which is better overall, Mercury 2.5 or Mercury 3?

Mercury 2.5 has the higher overall point estimate. Conditional score ranges do not establish rank confidence.

### How many benchmarks does Mercury 2.5 cover on BenchLM?

Mercury 2.5 currently has 6 sourced benchmark scores on BenchLM.

### How many benchmarks does Mercury 3 cover on BenchLM?

Mercury 3 currently has 8 sourced benchmark scores on BenchLM.

## Related Comparisons

- [Mercury 2.5 vs Mercury 2.5 Preview](/compare/mercury-2-5-vs-mercury-2-5-preview)
- [Mercury 3 vs Mercury 2.5 Preview](/compare/mercury-2-5-preview-vs-mercury-3)
- [Mercury 2.5 vs Claude Opus 5.5](/compare/claude-opus-5-5-vs-mercury-2-5)
- [Mercury 3 vs Claude Opus 5.5](/compare/claude-opus-5-5-vs-mercury-3)

## Explore More

- [Mercury 2.5 profile](/models/mercury-2-5)
- [Mercury 3 profile](/models/mercury-3)
- [Compare Pricing](/llm-pricing)
- [Alternative Finder](/tools/alternative-finder)
- [LLM Selector](/tools/llm-selector)
- [Overall Rankings](/best/overall)