# DeepSeek-V3.1 vs Gemini 2.5 Pro

> Gemini 2.5 Pro is the stronger model overall, scoring 44.9 to 42.2 on the Noometry Index. DeepSeek-V3.1 costs 8.1× 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/deepseek-v3-1-vs-gemini-2-5-pro
- Last updated: 2026-10-10
- Shared benchmarks: 27

## Summary

- They share 27 benchmarks with published results for both. DeepSeek-V3.1 scores higher in 1 category and Gemini 2.5 Pro in 7 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in long context, where Gemini 2.5 Pro leads 59.7 to 36.3.
- The biggest single-benchmark swing is Fiction.LiveBench: 52.8% for DeepSeek-V3.1 and 91.7% for Gemini 2.5 Pro.
- DeepSeek-V3.1 is cheaper at $0.25 / $0.95 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 164K.
- DeepSeek-V3.1 has downloadable open weights; the other is API-only.

## Snapshot

| | DeepSeek-V3.1 | Gemini 2.5 Pro |
|---|---|---|
| Provider | DeepSeek | Google |
| Noometry Index | 42.2 | 44.9 |
| Rank | 106 | 67 |
| Context | 164K | 1.05M |
| Input $/M | $0.25 | $1.25 |
| Output $/M | $0.95 | $10 |
| Weights | Open | Proprietary |

## Coding

- DeepSeek-V3.1: 39.8 (#140)
- Gemini 2.5 Pro: 42.8 (#86)

| Benchmark | DeepSeek-V3.1 | Gemini 2.5 Pro |
|---|---|---|
| WeirdML | 38.4% | 54% |
| LMArena Coding | 1417 | 1452 |
| SWE-bench Verified | — | 57.6% |
| SWE-bench Verified (bash only) | — | 53.6% |
| Aider Polyglot | — | 83.1% |
| LMArena WebDev | — | 1227 |
| SciCode | — | 42.8% |
| GSO | — | 3.9% |
| LiveBench Coding | — | 85.9% |
| CadEval | — | 64% |
| ALE-Bench | — | 785.52 |
| AlgoTune | — | 1.51 |

## Agentic & Tool Use

- DeepSeek-V3.1: —
- Gemini 2.5 Pro: 29.1 (#81)

| Benchmark | DeepSeek-V3.1 | Gemini 2.5 Pro |
|---|---|---|
| Terminal-Bench | — | 32.6% |
| GDPval | — | 23.3% |
| Remote Labor Index | — | 0.8% |
| TheAgentCompany | — | 30.3% |
| τ²-bench Banking | — | 13.7% |
| DeepResearch Bench | — | 42.8% |
| BALROG | — | 43.3% |
| LMArena Search | — | 1142 |
| METR Time Horizons | — | 55.4% |
| Vending-Bench 2 | — | 573.64 |

## Reasoning

- DeepSeek-V3.1: 27.3 (#109)
- Gemini 2.5 Pro: 28.8 (#85)

| Benchmark | DeepSeek-V3.1 | Gemini 2.5 Pro |
|---|---|---|
| SimpleBench | 40% | 62.4% |
| Kagi LLM Benchmark | 53.2% | 70.3% |
| LMArena Hard Prompts | 1417 | 1455 |
| DTBench | 82.7% | 82.4% |
| LMCA | 24.3% | 34.8% |
| Epoch Capabilities Index | 139.92 | 145.32 |
| ForecastBench | 58 | 61.3 |
| ARC-AGI-2 | — | 4.9% |
| ARC-AGI-1 | — | 41% |
| CritPt | — | 2% |
| Chess Puzzles | — | 20% |
| EnigmaEval | — | 5.6% |
| LiveBench Reasoning | — | 89.8% |
| LiveBench Data Analysis | — | 79.9% |
| LiveBench | — | 82.3% |

## Math

- DeepSeek-V3.1: 36.1 (#125)
- Gemini 2.5 Pro: 31.9 (#191)

| Benchmark | DeepSeek-V3.1 | Gemini 2.5 Pro |
|---|---|---|
| LMArena Math | 1420 | 1450 |
| FrontierMath (Tiers 1-3) | — | 24.6% |
| FrontierMath Tier 4 | — | 0% |
| OTIS Mock AIME 2024-2025 | — | 84.7% |
| Omni-MATH | — | 41.6% |
| LiveBench Math | — | 90.2% |
| MATH Level 5 | — | 95.9% |
| FrontierMath (Feb 2025 set) | — | 14.1% |
| FrontierMath Tier 4 (v1) | — | 4.2% |

## Knowledge

- DeepSeek-V3.1: 43.0 (#86)
- Gemini 2.5 Pro: 55.8 (#41)

| Benchmark | DeepSeek-V3.1 | Gemini 2.5 Pro |
|---|---|---|
| Vectara Hallucination Rate | 5.5% | 7% |
| LMArena Expert | 1403 | 1452 |
| GPQA Diamond | — | 85.3% |
| Humanity's Last Exam | — | 21.6% |
| MMLU-Pro | — | 86.3% |
| Confabulations | — | 10.6% |
| GPQA (HELM) | — | 74.9% |

## Multimodal

- DeepSeek-V3.1: —
- Gemini 2.5 Pro: 44.5 (#16)

| Benchmark | DeepSeek-V3.1 | Gemini 2.5 Pro |
|---|---|---|
| LMArena Vision | — | 1263 |
| GeoBench | — | 86% |
| VPCT | — | 48% |
| LMArena Document | — | 1421 |
| SpatialViz-Bench | — | 44.7% |

## Multilingual

- DeepSeek-V3.1: 51.8 (#100)
- Gemini 2.5 Pro: 55.5 (#30)

| Benchmark | DeepSeek-V3.1 | Gemini 2.5 Pro |
|---|---|---|
| LMArena Non-English | 1400 | 1452 |
| LMArena Chinese | 1470 | 1507 |
| LMArena French | 1447 | 1473 |
| LMArena German | 1412 | 1487 |
| LMArena Japanese | 1374 | 1462 |
| LMArena Korean | 1337 | 1435 |
| LMArena Russian | 1405 | 1460 |
| LMArena Spanish | 1429 | 1473 |

## Instruction Following

- DeepSeek-V3.1: 74.1 (#105)
- Gemini 2.5 Pro: 75.2 (#71)

| Benchmark | DeepSeek-V3.1 | Gemini 2.5 Pro |
|---|---|---|
| LMArena Instruction Following | 1400 | 1436 |
| LiveBench Instruction Following | — | 80.6% |
| IFEval | — | 84% |

## Long Context

- DeepSeek-V3.1: 36.3 (#227)
- Gemini 2.5 Pro: 59.7 (#5)

| Benchmark | DeepSeek-V3.1 | Gemini 2.5 Pro |
|---|---|---|
| Fiction.LiveBench | 52.8% | 91.7% |
| LMArena Longer Query | 1422 | 1449 |

## Writing & Preference

- DeepSeek-V3.1: 60.5 (#93)
- Gemini 2.5 Pro: 63.9 (#60)

| Benchmark | DeepSeek-V3.1 | Gemini 2.5 Pro |
|---|---|---|
| LMArena Text | 1420 | 1458 |
| LMArena Creative Writing | 1401 | 1453 |
| EQ-Bench Creative Writing | 1436 | 1421 |
| LMArena Multi-Turn | 1408 | 1453 |
| Short-Story Creative Writing | — | 83.8% |
| WildBench | — | 85.7% |
| LiveBench Language | — | 67.8% |

## FAQ

### Is DeepSeek-V3.1 better than Gemini 2.5 Pro?

Gemini 2.5 Pro is the stronger model overall, scoring 44.9 to 42.2 on the Noometry Index. DeepSeek-V3.1 costs 8.1× 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, DeepSeek-V3.1 or Gemini 2.5 Pro?

DeepSeek-V3.1 is cheaper. It lists at $0.25 per million input tokens and $0.95 per million output tokens; Gemini 2.5 Pro lists at $1.25 and $10.

### Is DeepSeek-V3.1 or Gemini 2.5 Pro better for coding?

Gemini 2.5 Pro scores higher on coding benchmarks: 42.8 versus 39.8 in the Noometry coding category.

### Which has the bigger context window?

Gemini 2.5 Pro does, with 1.05M tokens against 164K.

### How many benchmarks do DeepSeek-V3.1 and Gemini 2.5 Pro share?

27 benchmarks have published results for both models. DeepSeek-V3.1 has 27 scored results on Noometry and Gemini 2.5 Pro has 78.