# DeepSeek-V3.2-Exp vs Gemini 3.1 Pro Preview

> Gemini 3.1 Pro Preview is the stronger model overall, scoring 56.7 to 44.9 on the Noometry Index. DeepSeek-V3.2-Exp costs 16× less per token, which makes it the better buy when Gemini 3.1 Pro Preview's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/deepseek-v3-2-exp-vs-gemini-3-1-pro-preview
- Last updated: 2026-10-10
- Shared benchmarks: 41

## Summary

- They share 41 benchmarks with published results for both. DeepSeek-V3.2-Exp scores higher in 3 categories and Gemini 3.1 Pro Preview in 6 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Gemini 3.1 Pro Preview leads 71.7 to 22.0.
- The biggest single-benchmark swing is ARC-AGI-2: 4% for DeepSeek-V3.2-Exp and 77.1% for Gemini 3.1 Pro Preview.
- DeepSeek-V3.2-Exp is cheaper at $0.26 / $0.38 per million input/output tokens, against $2 / $12 for Gemini 3.1 Pro Preview.
- Gemini 3.1 Pro Preview accepts more context: 1.05M tokens versus 164K.
- DeepSeek-V3.2-Exp has downloadable open weights; the other is API-only.

## Snapshot

| | DeepSeek-V3.2-Exp | Gemini 3.1 Pro Preview |
|---|---|---|
| Provider | DeepSeek | Google |
| Noometry Index | 44.9 | 56.7 |
| Rank | 68 | 20 |
| Context | 164K | 1.05M |
| Input $/M | $0.26 | $2 |
| Output $/M | $0.38 | $12 |
| Weights | Open | Proprietary |

## Coding

- DeepSeek-V3.2-Exp: 46.4 (#61)
- Gemini 3.1 Pro Preview: 42.7 (#88)

| Benchmark | DeepSeek-V3.2-Exp | Gemini 3.1 Pro Preview |
|---|---|---|
| LMArena WebDev | 1362 | 1446 |
| SciCode | 38.9% | 58.9% |
| WeirdML | 39.5% | 72.1% |
| LMArena Coding | 1454 | 1483 |
| SWE-bench Verified | — | 75.6% |
| DeepSWE | — | 11.7% |
| SWE-bench Verified (bash only) | 70% | — |
| Aider Polyglot | 74.2% | — |
| SWE-bench Multilingual | 59% | — |
| GSO | — | 22.6% |
| MirrorCode | — | 8.9% |
| ALE-Bench | — | 1,161 |
| AlgoTune | — | 2.02 |

## Agentic & Tool Use

- DeepSeek-V3.2-Exp: 37.9 (#31)
- Gemini 3.1 Pro Preview: 37.7 (#33)

| Benchmark | DeepSeek-V3.2-Exp | Gemini 3.1 Pro Preview |
|---|---|---|
| Terminal-Bench | 39.6% | 80.2% |
| Vending-Bench 2 | 1,034 | 3,774 |
| APEX-Agents | — | 35.3% |
| Berkeley Function Calling Leaderboard | 56.7% | — |
| TheAgentCompany | 42.9% | — |
| τ²-bench Banking | — | 26% |
| DeepResearch Bench | — | 47.8% |
| PostTrainBench | — | 22% |
| BALROG | — | 57% |
| ExploitBench | — | 26.1% |
| GBAEval | — | 0.8% |
| GDP.pdf | — | 17% |
| LMArena Search | — | 1211 |
| METR Time Horizons | — | 77% |

## Reasoning

- DeepSeek-V3.2-Exp: 22.0 (#185)
- Gemini 3.1 Pro Preview: 71.7 (#9)

| Benchmark | DeepSeek-V3.2-Exp | Gemini 3.1 Pro Preview |
|---|---|---|
| ARC-AGI-2 | 4% | 77.1% |
| NYT Connections (extended) | 36.7% | 97.4% |
| ARC-AGI-1 | 57% | 98% |
| CritPt | 2.9% | 17.7% |
| Chess Puzzles | 14% | 55% |
| Thematic Generalization | 65% | 79.4% |
| LMArena Hard Prompts | 1433 | 1484 |
| DTBench | 87.7% | 97.1% |
| LMCA | 29.1% | 53.8% |
| Epoch Capabilities Index | 146.28 | 154.92 |
| SimpleBench | — | 79.6% |
| Kagi LLM Benchmark | 52.2% | — |
| EnigmaEval | — | 36.8% |
| EBR-Bench | — | 14.3% |
| Mystery Game Puzzles | — | 34% |
| ForecastBench | — | 59 |

## Math

- DeepSeek-V3.2-Exp: 40.8 (#75)
- Gemini 3.1 Pro Preview: 61.8 (#28)

| Benchmark | DeepSeek-V3.2-Exp | Gemini 3.1 Pro Preview |
|---|---|---|
| MathArena Final-Answer Competitions | 57.7% | 86.5% |
| OTIS Mock AIME 2024-2025 | 87.8% | 95.6% |
| ProofBench | 8% | 26% |
| LMArena Math | 1435 | 1485 |
| FrontierMath (Feb 2025 set) | 22.1% | 36.9% |
| FrontierMath Tier 4 (v1) | 2.1% | 16.7% |
| FrontierMath (Tiers 1-3) | — | 59.6% |
| FrontierMath Tier 4 | — | 26.8% |

## Knowledge

- DeepSeek-V3.2-Exp: 51.6 (#61)
- Gemini 3.1 Pro Preview: 71.8 (#3)

| Benchmark | DeepSeek-V3.2-Exp | Gemini 3.1 Pro Preview |
|---|---|---|
| GPQA Diamond | 83.4% | 94.4% |
| Vectara Hallucination Rate | 5.3% | 10.4% |
| LMArena Expert | 1436 | 1484 |
| Humanity's Last Exam | — | 46.4% |
| SimpleQA Verified | — | 73.5% |

## Multimodal

- DeepSeek-V3.2-Exp: —
- Gemini 3.1 Pro Preview: 38.1 (#60)

| Benchmark | DeepSeek-V3.2-Exp | Gemini 3.1 Pro Preview |
|---|---|---|
| LMArena Vision | — | 1296 |
| Blueprint-Bench 2 | — | 26.5% |
| Furniture Assembly | — | 26.7% |
| LMArena Document | — | 1444 |

## Multilingual

- DeepSeek-V3.2-Exp: 52.4 (#85)
- Gemini 3.1 Pro Preview: 57.1 (#11)

| Benchmark | DeepSeek-V3.2-Exp | Gemini 3.1 Pro Preview |
|---|---|---|
| LMArena Non-English | 1409 | 1475 |
| LMArena Chinese | 1461 | 1528 |
| LMArena French | 1435 | 1488 |
| LMArena German | 1440 | 1491 |
| LMArena Japanese | 1377 | 1494 |
| LMArena Korean | 1372 | 1458 |
| LMArena Russian | 1423 | 1496 |
| LMArena Spanish | 1441 | 1480 |

## Instruction Following

- DeepSeek-V3.2-Exp: 74.7 (#89)
- Gemini 3.1 Pro Preview: 77.1 (#31)

| Benchmark | DeepSeek-V3.2-Exp | Gemini 3.1 Pro Preview |
|---|---|---|
| LMArena Instruction Following | 1413 | 1465 |

## Long Context

- DeepSeek-V3.2-Exp: 47.6 (#15)
- Gemini 3.1 Pro Preview: 47.3 (#17)

| Benchmark | DeepSeek-V3.2-Exp | Gemini 3.1 Pro Preview |
|---|---|---|
| CL-bench | 13.2% | 20.8% |
| CL-bench Life | 9.5% | 16.9% |
| LMArena Longer Query | 1428 | 1482 |
| Fiction.LiveBench | 83.3% | — |

## Writing & Preference

- DeepSeek-V3.2-Exp: 62.5 (#73)
- Gemini 3.1 Pro Preview: 66.3 (#36)

| Benchmark | DeepSeek-V3.2-Exp | Gemini 3.1 Pro Preview |
|---|---|---|
| LMArena Text | 1425 | 1480 |
| LMArena Creative Writing | 1403 | 1482 |
| EQ-Bench Creative Writing | 1515 | 1491 |
| LMArena Multi-Turn | 1427 | 1486 |
| EQ-Bench 4 | — | 1142 |

## FAQ

### Is DeepSeek-V3.2-Exp better than Gemini 3.1 Pro Preview?

Gemini 3.1 Pro Preview is the stronger model overall, scoring 56.7 to 44.9 on the Noometry Index. DeepSeek-V3.2-Exp costs 16× less per token, which makes it the better buy when Gemini 3.1 Pro Preview's lead doesn't matter for your workload.

### Which is cheaper, DeepSeek-V3.2-Exp or Gemini 3.1 Pro Preview?

DeepSeek-V3.2-Exp is cheaper. It lists at $0.26 per million input tokens and $0.38 per million output tokens; Gemini 3.1 Pro Preview lists at $2 and $12.

### Is DeepSeek-V3.2-Exp or Gemini 3.1 Pro Preview better for coding?

DeepSeek-V3.2-Exp scores higher on coding benchmarks: 46.4 versus 42.7 in the Noometry coding category.

### Which has the bigger context window?

Gemini 3.1 Pro Preview does, with 1.05M tokens against 164K.

### How many benchmarks do DeepSeek-V3.2-Exp and Gemini 3.1 Pro Preview share?

41 benchmarks have published results for both models. DeepSeek-V3.2-Exp has 48 scored results on Noometry and Gemini 3.1 Pro Preview has 71.