# DeepSeek-V3.1 vs Gemini 3 Flash Preview

> Gemini 3 Flash Preview is the stronger model overall, scoring 52.3 to 42.2 on the Noometry Index. DeepSeek-V3.1 costs 2.6× less per token, which makes it the better buy when Gemini 3 Flash Preview's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/deepseek-v3-1-vs-gemini-3-flash-preview
- Last updated: 2026-10-10
- Shared benchmarks: 24

## Summary

- They share 24 benchmarks with published results for both. DeepSeek-V3.1 scores higher in 0 categories and Gemini 3 Flash Preview in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Gemini 3 Flash Preview leads 49.2 to 27.3.
- The biggest single-benchmark swing is WeirdML: 38.4% for DeepSeek-V3.1 and 61.6% for Gemini 3 Flash Preview.
- DeepSeek-V3.1 is cheaper at $0.25 / $0.95 per million input/output tokens, against $0.50 / $3 for Gemini 3 Flash Preview.
- Gemini 3 Flash Preview 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 3 Flash Preview |
|---|---|---|
| Provider | DeepSeek | Google |
| Noometry Index | 42.2 | 52.3 |
| Rank | 106 | 33 |
| Context | 164K | 1.05M |
| Input $/M | $0.25 | $0.50 |
| Output $/M | $0.95 | $3 |
| Weights | Open | Proprietary |

## Coding

- DeepSeek-V3.1: 39.8 (#140)
- Gemini 3 Flash Preview: 51.5 (#35)

| Benchmark | DeepSeek-V3.1 | Gemini 3 Flash Preview |
|---|---|---|
| WeirdML | 38.4% | 61.6% |
| LMArena Coding | 1417 | 1460 |
| SWE-bench Verified | — | 75.4% |
| SWE-bench Verified (bash only) | — | 75.8% |
| LMArena WebDev | — | 1439 |
| SWE-bench Multilingual | — | 72.7% |
| GSO | — | 9.8% |
| ALE-Bench | — | 1,367 |

## Agentic & Tool Use

- DeepSeek-V3.1: —
- Gemini 3 Flash Preview: 38.6 (#27)

| Benchmark | DeepSeek-V3.1 | Gemini 3 Flash Preview |
|---|---|---|
| Terminal-Bench | — | 64.3% |
| τ²-bench Airline | — | 82.5% |
| τ²-bench Banking | — | 27.3% |
| τ²-bench Retail | — | 76.8% |
| τ²-bench Telecom | — | 91.2% |
| DeepResearch Bench | — | 49.8% |
| BALROG | — | 48.1% |
| GDP.pdf | — | 10% |
| LMArena Search | — | 1198 |
| Vending-Bench 2 | — | 3,635 |

## Reasoning

- DeepSeek-V3.1: 27.3 (#109)
- Gemini 3 Flash Preview: 49.2 (#35)

| Benchmark | DeepSeek-V3.1 | Gemini 3 Flash Preview |
|---|---|---|
| SimpleBench | 40% | 61.1% |
| LMArena Hard Prompts | 1417 | 1465 |
| DTBench | 82.7% | 89.1% |
| LMCA | 24.3% | 43.1% |
| Epoch Capabilities Index | 139.92 | 151.83 |
| ForecastBench | 58 | 58.5 |
| ARC-AGI-2 | — | 33.6% |
| Kagi LLM Benchmark | 53.2% | — |
| NYT Connections (extended) | — | 83.1% |
| ARC-AGI-1 | — | 84.7% |
| Chess Puzzles | — | 40% |
| Mystery Game Puzzles | — | 26% |

## Math

- DeepSeek-V3.1: 36.1 (#125)
- Gemini 3 Flash Preview: 51.4 (#42)

| Benchmark | DeepSeek-V3.1 | Gemini 3 Flash Preview |
|---|---|---|
| LMArena Math | 1420 | 1472 |
| FrontierMath (Tiers 1-3) | — | 51.2% |
| FrontierMath Tier 4 | — | 17.1% |
| MathArena Final-Answer Competitions | — | 67.6% |
| OTIS Mock AIME 2024-2025 | — | 95.6% |
| ProofBench | — | 15% |
| FrontierMath (Feb 2025 set) | — | 35.6% |
| FrontierMath Tier 4 (v1) | — | 4.2% |

## Knowledge

- DeepSeek-V3.1: 43.0 (#86)
- Gemini 3 Flash Preview: 58.8 (#30)

| Benchmark | DeepSeek-V3.1 | Gemini 3 Flash Preview |
|---|---|---|
| Vectara Hallucination Rate | 5.5% | 13.5% |
| LMArena Expert | 1403 | 1461 |
| GPQA Diamond | — | 89.4% |
| SimpleQA Verified | — | 66.8% |

## Multimodal

- DeepSeek-V3.1: —
- Gemini 3 Flash Preview: 45.3 (#15)

| Benchmark | DeepSeek-V3.1 | Gemini 3 Flash Preview |
|---|---|---|
| LMArena Vision | — | 1285 |
| GeoBench | — | 88% |
| VPCT | — | 72.6% |
| Blueprint-Bench 2 | — | 0% |
| LMArena Document | — | 1413 |

## Multilingual

- DeepSeek-V3.1: 51.8 (#100)
- Gemini 3 Flash Preview: 55.9 (#26)

| Benchmark | DeepSeek-V3.1 | Gemini 3 Flash Preview |
|---|---|---|
| LMArena Non-English | 1400 | 1458 |
| LMArena Chinese | 1470 | 1510 |
| LMArena French | 1447 | 1478 |
| LMArena German | 1412 | 1497 |
| LMArena Japanese | 1374 | 1488 |
| LMArena Korean | 1337 | 1443 |
| LMArena Russian | 1405 | 1480 |
| LMArena Spanish | 1429 | 1470 |

## Instruction Following

- DeepSeek-V3.1: 74.1 (#105)
- Gemini 3 Flash Preview: 75.8 (#52)

| Benchmark | DeepSeek-V3.1 | Gemini 3 Flash Preview |
|---|---|---|
| LMArena Instruction Following | 1400 | 1437 |

## Long Context

- DeepSeek-V3.1: 36.3 (#227)
- Gemini 3 Flash Preview: 44.4 (#63)

| Benchmark | DeepSeek-V3.1 | Gemini 3 Flash Preview |
|---|---|---|
| LMArena Longer Query | 1422 | 1452 |
| Fiction.LiveBench | 52.8% | — |

## Writing & Preference

- DeepSeek-V3.1: 60.5 (#93)
- Gemini 3 Flash Preview: 65.6 (#44)

| Benchmark | DeepSeek-V3.1 | Gemini 3 Flash Preview |
|---|---|---|
| LMArena Text | 1420 | 1466 |
| LMArena Creative Writing | 1401 | 1457 |
| LMArena Multi-Turn | 1408 | 1471 |
| EQ-Bench Creative Writing | 1436 | — |

## FAQ

### Is DeepSeek-V3.1 better than Gemini 3 Flash Preview?

Gemini 3 Flash Preview is the stronger model overall, scoring 52.3 to 42.2 on the Noometry Index. DeepSeek-V3.1 costs 2.6× less per token, which makes it the better buy when Gemini 3 Flash Preview's lead doesn't matter for your workload.

### Which is cheaper, DeepSeek-V3.1 or Gemini 3 Flash Preview?

DeepSeek-V3.1 is cheaper. It lists at $0.25 per million input tokens and $0.95 per million output tokens; Gemini 3 Flash Preview lists at $0.50 and $3.

### Is DeepSeek-V3.1 or Gemini 3 Flash Preview better for coding?

Gemini 3 Flash Preview scores higher on coding benchmarks: 51.5 versus 39.8 in the Noometry coding category.

### Which has the bigger context window?

Gemini 3 Flash Preview does, with 1.05M tokens against 164K.

### How many benchmarks do DeepSeek-V3.1 and Gemini 3 Flash Preview share?

24 benchmarks have published results for both models. DeepSeek-V3.1 has 27 scored results on Noometry and Gemini 3 Flash Preview has 59.