# DeepSeek-V3 vs Gemini 3.5 Flash Lite

> Gemini 3.5 Flash Lite is the stronger model overall, scoring 41.5 to 39.5 on the Noometry Index. DeepSeek-V3 costs 2.1× less per token, which makes it the better buy when Gemini 3.5 Flash Lite's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/deepseek-v3-vs-gemini-3-5-flash-lite
- Last updated: 2026-10-10
- Shared benchmarks: 26

## Summary

- They share 26 benchmarks with published results for both. DeepSeek-V3 scores higher in 1 category and Gemini 3.5 Flash Lite in 7 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Gemini 3.5 Flash Lite leads 50.1 to 37.5.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 37.8% for DeepSeek-V3 and 71.1% for Gemini 3.5 Flash Lite.
- DeepSeek-V3 is cheaper at $0.24 / $0.90 per million input/output tokens, against $0.30 / $2.50 for Gemini 3.5 Flash Lite.
- Gemini 3.5 Flash Lite accepts more context: 1.05M tokens versus 164K.
- DeepSeek-V3 has downloadable open weights; the other is API-only.

## Snapshot

| | DeepSeek-V3 | Gemini 3.5 Flash Lite |
|---|---|---|
| Provider | DeepSeek | Google |
| Noometry Index | 39.5 | 41.5 |
| Rank | 166 | 133 |
| Context | 164K | 1.05M |
| Input $/M | $0.24 | $0.30 |
| Output $/M | $0.90 | $2.50 |
| Weights | Open | Proprietary |

## Coding

- DeepSeek-V3: 42.3 (#106)
- Gemini 3.5 Flash Lite: 42.4 (#104)

| Benchmark | DeepSeek-V3 | Gemini 3.5 Flash Lite |
|---|---|---|
| SciCode | 35.8% | 41.3% |
| WeirdML | 36.1% | 39% |
| LMArena Coding | 1368 | 1453 |
| Aider Polyglot | 55.1% | — |
| LMArena WebDev | — | 1440 |
| BigCodeBench Instruct | 50% | — |
| LiveBench Coding | 70.9% | — |
| BigCodeBench Complete | 62.2% | — |
| ALE-Bench | — | 765.27 |
| HumanEval+ | 86.6% | — |
| MBPP+ | 73% | — |

## Agentic & Tool Use

- DeepSeek-V3: —
- Gemini 3.5 Flash Lite: 25.7 (#107)

| Benchmark | DeepSeek-V3 | Gemini 3.5 Flash Lite |
|---|---|---|
| APEX-Agents | — | 29.3% |
| GDP.pdf | — | 10% |
| METR Time Horizons | 49.6% | — |

## Reasoning

- DeepSeek-V3: 20.5 (#236)
- Gemini 3.5 Flash Lite: 27.8 (#114)

| Benchmark | DeepSeek-V3 | Gemini 3.5 Flash Lite |
|---|---|---|
| CritPt | 0% | 0% |
| LMArena Hard Prompts | 1365 | 1441 |
| DTBench | 64.8% | 83.5% |
| LMCA | 15.5% | 37.6% |
| Epoch Capabilities Index | 135.94 | 145.13 |
| ARC-AGI-2 | — | 10.3% |
| SimpleBench | 27.2% | — |
| Kagi LLM Benchmark | 52.3% | — |
| NYT Connections (extended) | — | 60.4% |
| ARC-AGI-1 | — | 53.5% |
| Chess Puzzles | — | 22% |
| LiveBench Reasoning | 65.8% | — |
| Mystery Game Puzzles | — | 19% |
| LiveBench Data Analysis | 60.9% | — |
| BIG-Bench Hard | 87.5% | — |
| ForecastBench | 59.1 | — |
| HellaSwag | 88.9% | — |
| LiveBench | 66.9% | — |
| PIQA | 84.7% | — |
| WinoGrande | 85.2% | — |

## Math

- DeepSeek-V3: 32.1 (#219)
- Gemini 3.5 Flash Lite: 25.9 (#264)

| Benchmark | DeepSeek-V3 | Gemini 3.5 Flash Lite |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 37.8% | 71.1% |
| LMArena Math | 1373 | 1433 |
| FrontierMath (Tiers 1-3) | — | 26% |
| FrontierMath Tier 4 | — | 0% |
| ProofBench | — | 13% |
| Omni-MATH | 40.3% | — |
| LiveBench Math | 73.5% | — |
| MATH Level 5 | 75.5% | — |
| FrontierMath (Feb 2025 set) | 1.7% | — |

## Knowledge

- DeepSeek-V3: 37.5 (#155)
- Gemini 3.5 Flash Lite: 50.1 (#72)

| Benchmark | DeepSeek-V3 | Gemini 3.5 Flash Lite |
|---|---|---|
| GPQA Diamond | 67.6% | 83.3% |
| LMArena Expert | 1351 | 1435 |
| MMLU-Pro | 72.3% | — |
| Confabulations | 26.1% | — |
| Vectara Hallucination Rate | 6.1% | — |
| GPQA (HELM) | 53.8% | — |
| ARC (AI2) Challenge | 95.3% | — |
| MMLU | 87.2% | — |
| TriviaQA | 82.9% | — |

## Multimodal

- DeepSeek-V3: —
- Gemini 3.5 Flash Lite: 41.1 (#40)

| Benchmark | DeepSeek-V3 | Gemini 3.5 Flash Lite |
|---|---|---|
| LMArena Vision | — | 1268 |

## Multilingual

- DeepSeek-V3: 48.5 (#143)
- Gemini 3.5 Flash Lite: 53.5 (#63)

| Benchmark | DeepSeek-V3 | Gemini 3.5 Flash Lite |
|---|---|---|
| LMArena Non-English | 1358 | 1427 |
| LMArena Chinese | 1391 | 1469 |
| LMArena French | 1385 | 1451 |
| LMArena German | 1374 | 1454 |
| LMArena Japanese | 1333 | 1427 |
| LMArena Korean | 1319 | 1407 |
| LMArena Russian | 1373 | 1443 |
| LMArena Spanish | 1358 | 1442 |

## Instruction Following

- DeepSeek-V3: 72.8 (#130)
- Gemini 3.5 Flash Lite: 74.8 (#85)

| Benchmark | DeepSeek-V3 | Gemini 3.5 Flash Lite |
|---|---|---|
| LMArena Instruction Following | 1345 | 1420 |
| LiveBench Instruction Following | 81.5% | — |
| IFEval | 83.2% | — |

## Long Context

- DeepSeek-V3: 34.0 (#253)
- Gemini 3.5 Flash Lite: 43.9 (#84)

| Benchmark | DeepSeek-V3 | Gemini 3.5 Flash Lite |
|---|---|---|
| LMArena Longer Query | 1352 | 1435 |
| Fiction.LiveBench | 50% | — |

## Writing & Preference

- DeepSeek-V3: 57.4 (#130)
- Gemini 3.5 Flash Lite: 64.2 (#57)

| Benchmark | DeepSeek-V3 | Gemini 3.5 Flash Lite |
|---|---|---|
| LMArena Text | 1375 | 1435 |
| LMArena Creative Writing | 1364 | 1420 |
| EQ-Bench Creative Writing | 1472 | 1559 |
| LMArena Multi-Turn | 1389 | 1445 |
| Short-Story Creative Writing | 77% | — |
| WildBench | 83% | — |
| LiveBench Language | 49.1% | — |

## FAQ

### Is DeepSeek-V3 better than Gemini 3.5 Flash Lite?

Gemini 3.5 Flash Lite is the stronger model overall, scoring 41.5 to 39.5 on the Noometry Index. DeepSeek-V3 costs 2.1× less per token, which makes it the better buy when Gemini 3.5 Flash Lite's lead doesn't matter for your workload.

### Which is cheaper, DeepSeek-V3 or Gemini 3.5 Flash Lite?

DeepSeek-V3 is cheaper. It lists at $0.24 per million input tokens and $0.90 per million output tokens; Gemini 3.5 Flash Lite lists at $0.30 and $2.50.

### Is DeepSeek-V3 or Gemini 3.5 Flash Lite better for coding?

They score almost the same on coding (42.3 vs 42.4); test both on your own repository before choosing.

### Which has the bigger context window?

Gemini 3.5 Flash Lite does, with 1.05M tokens against 164K.

### How many benchmarks do DeepSeek-V3 and Gemini 3.5 Flash Lite share?

26 benchmarks have published results for both models. DeepSeek-V3 has 60 scored results on Noometry and Gemini 3.5 Flash Lite has 39.