# DeepSeek-V2.5 (Sep 2024) vs Gemini 2.0 Flash (Feb 2025)

> DeepSeek-V2.5 (Sep 2024) is the stronger model overall, scoring 37.0 to 34.8 on the Noometry Index.

- Canonical page: https://noometry.com/compare/deepseek-v2-5-vs-gemini-2-0-flash
- Last updated: 2026-10-10
- Shared benchmarks: 20

## Summary

- They share 20 benchmarks with published results for both. DeepSeek-V2.5 (Sep 2024) scores higher in 5 categories and Gemini 2.0 Flash (Feb 2025) in 3 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where DeepSeek-V2.5 (Sep 2024) leads 24.5 to 15.2.
- The biggest single-benchmark swing is Aider Polyglot: 17.8% for DeepSeek-V2.5 (Sep 2024) and 38.2% for Gemini 2.0 Flash (Feb 2025).
- DeepSeek-V2.5 (Sep 2024) has downloadable open weights; the other is API-only.

## Snapshot

| | DeepSeek-V2.5 (Sep 2024) | Gemini 2.0 Flash (Feb 2025) |
|---|---|---|
| Provider | DeepSeek | Google |
| Noometry Index | 37.0 | 34.8 |
| Rank | 197 | 223 |
| Context | — | — |
| Input $/M | — | — |
| Output $/M | — | — |
| Weights | Open | Proprietary |

## Coding

- DeepSeek-V2.5 (Sep 2024): 31.3 (#274)
- Gemini 2.0 Flash (Feb 2025): 28.2 (#308)

| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 2.0 Flash (Feb 2025) |
|---|---|---|
| Aider Polyglot | 17.8% | 38.2% |
| BigCodeBench Instruct | 48.6% | 45.9% |
| LMArena Coding | 1310 | 1351 |
| BigCodeBench Complete | 53.2% | 59.9% |
| SWE-bench Verified (bash only) | — | 13.5% |
| WeirdML | — | 25.8% |
| LiveBench Coding | — | 63.4% |
| CadEval | — | 30% |
| HumanEval+ | 83.5% | — |
| MBPP+ | 74.1% | — |

## Agentic & Tool Use

- DeepSeek-V2.5 (Sep 2024): —
- Gemini 2.0 Flash (Feb 2025): 28.0 (#85)

| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 2.0 Flash (Feb 2025) |
|---|---|---|
| TheAgentCompany | — | 11.4% |

## Reasoning

- DeepSeek-V2.5 (Sep 2024): 24.5 (#145)
- Gemini 2.0 Flash (Feb 2025): 15.2 (#308)

| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 2.0 Flash (Feb 2025) |
|---|---|---|
| LMArena Hard Prompts | 1289 | 1346 |
| ARC-AGI-2 | — | 1.3% |
| SimpleBench | — | 31.1% |
| Kagi LLM Benchmark | — | 37.8% |
| EnigmaEval | — | 1.1% |
| LiveBench Reasoning | — | 78.2% |
| DTBench | — | 63.2% |
| LiveBench Data Analysis | — | 69.4% |
| Epoch Capabilities Index | — | 135.36 |
| LiveBench | — | 66.9% |

## Math

- DeepSeek-V2.5 (Sep 2024): 33.3 (#177)
- Gemini 2.0 Flash (Feb 2025): 36.1 (#127)

| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 2.0 Flash (Feb 2025) |
|---|---|---|
| LMArena Math | 1289 | 1351 |
| OTIS Mock AIME 2024-2025 | — | 57.8% |
| Omni-MATH | — | 45.9% |
| LiveBench Math | — | 75.8% |
| MATH Level 5 | — | 82.2% |
| FrontierMath (Feb 2025 set) | — | 1.7% |

## Knowledge

- DeepSeek-V2.5 (Sep 2024): 34.2 (#190)
- Gemini 2.0 Flash (Feb 2025): 32.0 (#207)

| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 2.0 Flash (Feb 2025) |
|---|---|---|
| LMArena Expert | 1266 | 1340 |
| GPQA Diamond | — | 64.1% |
| Humanity's Last Exam | — | 6.6% |
| MMLU-Pro | — | 73.7% |
| Confabulations | — | 12.4% |
| GPQA (HELM) | — | 55.6% |
| MMLU | — | 79.7% |

## Multimodal

- DeepSeek-V2.5 (Sep 2024): —
- Gemini 2.0 Flash (Feb 2025): 35.9 (#74)

| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 2.0 Flash (Feb 2025) |
|---|---|---|
| LMArena Vision | — | 1158 |
| GeoBench | — | 77% |

## Multilingual

- DeepSeek-V2.5 (Sep 2024): 42.7 (#188)
- Gemini 2.0 Flash (Feb 2025): 47.6 (#144)

| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 2.0 Flash (Feb 2025) |
|---|---|---|
| LMArena Non-English | 1273 | 1342 |
| LMArena Chinese | 1317 | 1373 |
| LMArena French | 1288 | 1390 |
| LMArena German | 1258 | 1354 |
| LMArena Japanese | 1228 | 1293 |
| LMArena Korean | 1209 | 1314 |
| LMArena Russian | 1289 | 1351 |
| LMArena Spanish | 1248 | 1363 |

## Instruction Following

- DeepSeek-V2.5 (Sep 2024): 67.7 (#189)
- Gemini 2.0 Flash (Feb 2025): 74.6 (#91)

| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 2.0 Flash (Feb 2025) |
|---|---|---|
| LMArena Instruction Following | 1280 | 1336 |
| LiveBench Instruction Following | — | 85.8% |
| IFEval | — | 84.1% |

## Long Context

- DeepSeek-V2.5 (Sep 2024): 39.5 (#169)
- Gemini 2.0 Flash (Feb 2025): 38.1 (#198)

| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 2.0 Flash (Feb 2025) |
|---|---|---|
| LMArena Longer Query | 1301 | 1345 |
| Fiction.LiveBench | — | 61.1% |

## Writing & Preference

- DeepSeek-V2.5 (Sep 2024): 50.0 (#182)
- Gemini 2.0 Flash (Feb 2025): 49.8 (#184)

| Benchmark | DeepSeek-V2.5 (Sep 2024) | Gemini 2.0 Flash (Feb 2025) |
|---|---|---|
| LMArena Text | 1294 | 1354 |
| LMArena Creative Writing | 1285 | 1340 |
| LMArena Multi-Turn | 1297 | 1350 |
| Short-Story Creative Writing | — | 73.8% |
| EQ-Bench Creative Writing | — | 1128 |
| WildBench | — | 80% |
| LiveBench Language | — | 51.3% |

## FAQ

### Is DeepSeek-V2.5 (Sep 2024) better than Gemini 2.0 Flash (Feb 2025)?

DeepSeek-V2.5 (Sep 2024) is the stronger model overall, scoring 37.0 to 34.8 on the Noometry Index.

### Is DeepSeek-V2.5 (Sep 2024) or Gemini 2.0 Flash (Feb 2025) better for coding?

DeepSeek-V2.5 (Sep 2024) scores higher on coding benchmarks: 31.3 versus 28.2 in the Noometry coding category.

### How many benchmarks do DeepSeek-V2.5 (Sep 2024) and Gemini 2.0 Flash (Feb 2025) share?

20 benchmarks have published results for both models. DeepSeek-V2.5 (Sep 2024) has 22 scored results on Noometry and Gemini 2.0 Flash (Feb 2025) has 54.