# Claude Haiku 4.5 vs Gemini 2.5 Pro

> Gemini 2.5 Pro is the stronger model overall, scoring 45.0 to 39.5 on the Noometry Index. Claude Haiku 4.5 costs 1.7× less per token, which makes it the better buy when Gemini 2.5 Pro's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/claude-haiku-4-5-vs-gemini-2-5-pro
- Last updated: 2026-10-12
- Shared benchmarks: 46

## Summary

- They share 46 benchmarks with published results for both. Claude Haiku 4.5 scores higher in 3 categories and Gemini 2.5 Pro in 7 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in multimodal, where Gemini 2.5 Pro leads 45.2 to 26.8.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 66.7% for Claude Haiku 4.5 and 84.7% for Gemini 2.5 Pro.
- Claude Haiku 4.5 is cheaper at $1 / $5 per million input/output tokens, against $1.25 / $10 for Gemini 2.5 Pro.
- Gemini 2.5 Pro accepts more context: 1.05M tokens versus 200K.

## Snapshot

| | Claude Haiku 4.5 | Gemini 2.5 Pro |
|---|---|---|
| Provider | Anthropic | Google |
| Noometry Index | 39.5 | 45.0 |
| Rank | 165 | 75 |
| Context | 200K | 1.05M |
| Input $/M | $1 | $1.25 |
| Output $/M | $5 | $10 |
| Weights | Proprietary | Proprietary |

## Coding

- Claude Haiku 4.5: 44.0 (#78)
- Gemini 2.5 Pro: 42.4 (#101)

| Benchmark | Claude Haiku 4.5 | Gemini 2.5 Pro |
|---|---|---|
| SWE-bench Verified (bash only) | 66.6% | 53.6% |
| LMArena WebDev | 1330 | 1227 |
| SciCode | 43.3% | 42.8% |
| WeirdML | 45.4% | 54% |
| LMArena Coding | 1453 | 1452 |
| ALE-Bench | 653.48 | 785.52 |
| SWE-bench Verified | — | 57.6% |
| Aider Polyglot | — | 83.1% |
| SWE-bench Multilingual | 64.7% | — |
| GSO | — | 3.9% |
| LiveBench Coding | — | 85.9% |
| CadEval | — | 64% |
| AlgoTune | — | 1.51 |

## Agentic & Tool Use

- Claude Haiku 4.5: 33.6 (#52)
- Gemini 2.5 Pro: 29.2 (#88)

| Benchmark | Claude Haiku 4.5 | Gemini 2.5 Pro |
|---|---|---|
| Terminal-Bench | 35.5% | 32.6% |
| DeepResearch Bench | 45.5% | 42.8% |
| BALROG | 31.2% | 43.3% |
| Vending-Bench 2 | 458.89 | 573.64 |
| Berkeley Function Calling Leaderboard | 68.7% | — |
| GDPval | — | 23.3% |
| Remote Labor Index | — | 0.8% |
| TheAgentCompany | — | 30.3% |
| τ²-bench Banking | — | 13.7% |
| ExploitBench | 13.7% | — |
| LMArena Search | — | 1142 |
| METR Time Horizons | — | 55.4% |

## Reasoning

- Claude Haiku 4.5: 15.1 (#320)
- Gemini 2.5 Pro: 28.8 (#99)

| Benchmark | Claude Haiku 4.5 | Gemini 2.5 Pro |
|---|---|---|
| ARC-AGI-2 | 4% | 4.9% |
| ARC-AGI-1 | 47.7% | 41% |
| CritPt | 0% | 2% |
| Chess Puzzles | 8% | 20% |
| LMArena Hard Prompts | 1420 | 1455 |
| DTBench | 73.6% | 82.4% |
| LMCA | 30.9% | 34.8% |
| Epoch Capabilities Index | 142.41 | 145.32 |
| ForecastBench | 61.4 | 61.3 |
| SimpleBench | — | 62.4% |
| Kagi LLM Benchmark | — | 70.3% |
| NYT Connections (extended) | 14.3% | — |
| EnigmaEval | — | 5.6% |
| LiveBench Reasoning | — | 89.8% |
| LiveBench Data Analysis | — | 79.9% |
| LiveBench | — | 82.3% |

## Math

- Claude Haiku 4.5: 44.9 (#78)
- Gemini 2.5 Pro: 32.5 (#213)

| Benchmark | Claude Haiku 4.5 | Gemini 2.5 Pro |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 66.7% | 84.7% |
| Omni-MATH | 56.1% | 41.6% |
| LMArena Math | 1396 | 1450 |
| MATH Level 5 | 96.4% | 95.9% |
| FrontierMath (Feb 2025 set) | 5.9% | 14.1% |
| FrontierMath Tier 4 (v1) | 2.1% | 4.2% |
| FrontierMath (Tiers 1-3) | — | 24.6% |
| FrontierMath Tier 4 | — | 0% |
| LiveBench Math | — | 90.2% |

## Knowledge

- Claude Haiku 4.5: 37.7 (#153)
- Gemini 2.5 Pro: 56.0 (#46)

| Benchmark | Claude Haiku 4.5 | Gemini 2.5 Pro |
|---|---|---|
| GPQA Diamond | 71.2% | 85.3% |
| MMLU-Pro | 77.7% | 86.3% |
| Vectara Hallucination Rate | 9.8% | 7% |
| GPQA (HELM) | 60.5% | 74.9% |
| LMArena Expert | 1442 | 1452 |
| Humanity's Last Exam | — | 21.6% |
| SimpleQA Verified | 13.2% | — |
| Confabulations | — | 10.6% |

## Multimodal

- Claude Haiku 4.5: 26.8 (#118)
- Gemini 2.5 Pro: 45.2 (#18)

| Benchmark | Claude Haiku 4.5 | Gemini 2.5 Pro |
|---|---|---|
| LMArena Document | 1420 | 1421 |
| LMArena Vision | — | 1263 |
| GeoBench | — | 86% |
| VPCT | — | 48% |
| Blueprint-Bench 2 | 0% | — |
| SpatialViz-Bench | — | 44.7% |

## Multilingual

- Claude Haiku 4.5: 49.9 (#129)
- Gemini 2.5 Pro: 55.3 (#31)

| Benchmark | Claude Haiku 4.5 | Gemini 2.5 Pro |
|---|---|---|
| LMArena Non-English | 1377 | 1451 |
| LMArena Chinese | 1417 | 1507 |
| LMArena French | 1408 | 1472 |
| LMArena German | 1375 | 1487 |
| LMArena Japanese | 1339 | 1461 |
| LMArena Korean | 1347 | 1434 |
| LMArena Russian | 1381 | 1461 |
| LMArena Spanish | 1420 | 1473 |

## Instruction Following

- Claude Haiku 4.5: 71.4 (#149)
- Gemini 2.5 Pro: 75.0 (#75)

| Benchmark | Claude Haiku 4.5 | Gemini 2.5 Pro |
|---|---|---|
| IFEval | 80.1% | 84% |
| LMArena Instruction Following | 1414 | 1437 |
| LiveBench Instruction Following | — | 80.6% |

## Long Context

- Claude Haiku 4.5: 43.6 (#92)
- Gemini 2.5 Pro: 59.8 (#5)

| Benchmark | Claude Haiku 4.5 | Gemini 2.5 Pro |
|---|---|---|
| LMArena Longer Query | 1427 | 1449 |
| Fiction.LiveBench | — | 91.7% |

## Writing & Preference

- Claude Haiku 4.5: 57.9 (#123)
- Gemini 2.5 Pro: 63.7 (#62)

| Benchmark | Claude Haiku 4.5 | Gemini 2.5 Pro |
|---|---|---|
| LMArena Text | 1396 | 1458 |
| LMArena Creative Writing | 1372 | 1454 |
| WildBench | 83.9% | 85.7% |
| LMArena Multi-Turn | 1409 | 1453 |
| Short-Story Creative Writing | — | 83.8% |
| EQ-Bench Creative Writing | — | 1421 |
| EQ-Bench 4 | 1064 | — |
| LiveBench Language | — | 67.8% |

## FAQ

### Is Claude Haiku 4.5 better than Gemini 2.5 Pro?

Gemini 2.5 Pro is the stronger model overall, scoring 45.0 to 39.5 on the Noometry Index. Claude Haiku 4.5 costs 1.7× less per token, which makes it the better buy when Gemini 2.5 Pro's lead doesn't matter for your workload.

### Which is cheaper, Claude Haiku 4.5 or Gemini 2.5 Pro?

Claude Haiku 4.5 is cheaper. It lists at $1 per million input tokens and $5 per million output tokens; Gemini 2.5 Pro lists at $1.25 and $10.

### Is Claude Haiku 4.5 or Gemini 2.5 Pro better for coding?

Claude Haiku 4.5 scores higher on coding benchmarks: 44.0 versus 42.4 in the Noometry coding category.

### Which has the bigger context window?

Gemini 2.5 Pro does, with 1.05M tokens against 200K.

### How many benchmarks do Claude Haiku 4.5 and Gemini 2.5 Pro share?

46 benchmarks have published results for both models. Claude Haiku 4.5 has 53 scored results on Noometry and Gemini 2.5 Pro has 78.