# Gemini 3.1 Pro Preview vs o3

> Gemini 3.1 Pro Preview is the stronger model overall, scoring 56.7 to 47.5 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gemini-3-1-pro-preview-vs-o3
- Last updated: 2026-10-11
- Shared benchmarks: 45

## Summary

- They share 45 benchmarks with published results for both. Gemini 3.1 Pro Preview scores higher in 7 categories and o3 in 3 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Gemini 3.1 Pro Preview leads 71.7 to 32.0.
- The biggest single-benchmark swing is ARC-AGI-2: 77.1% for Gemini 3.1 Pro Preview and 6.5% for o3.
- o3 is cheaper at $2 / $8 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 200K.

## Snapshot

| | Gemini 3.1 Pro Preview | o3 |
|---|---|---|
| Provider | Google | OpenAI |
| Noometry Index | 56.7 | 47.5 |
| Rank | 23 | 61 |
| Context | 1.05M | 200K |
| Input $/M | $2 | $2 |
| Output $/M | $12 | $8 |
| Weights | Proprietary | Proprietary |

## Coding

- Gemini 3.1 Pro Preview: 42.5 (#99)
- o3: 46.8 (#64)

| Benchmark | Gemini 3.1 Pro Preview | o3 |
|---|---|---|
| SWE-bench Verified | 75.6% | 62.3% |
| GSO | 22.6% | 8.8% |
| WeirdML | 72.1% | 52.4% |
| LMArena Coding | 1484 | 1408 |
| ALE-Bench | 1,161 | 933.55 |
| DeepSWE | 11.7% | — |
| SWE-bench Verified (bash only) | — | 58.4% |
| Aider Polyglot | — | 81.3% |
| LMArena WebDev | 1447 | — |
| SciCode | 58.9% | — |
| MirrorCode | 8.9% | — |
| CadEval | — | 74% |
| AlgoTune | 2.02 | — |

## Agentic & Tool Use

- Gemini 3.1 Pro Preview: 37.7 (#34)
- o3: 34.5 (#44)

| Benchmark | Gemini 3.1 Pro Preview | o3 |
|---|---|---|
| DeepResearch Bench | 47.8% | 45.2% |
| LMArena Search | 1211 | 1144 |
| METR Time Horizons | 77% | 65.4% |
| Terminal-Bench | 80.2% | — |
| APEX-Agents | 35.3% | — |
| Berkeley Function Calling Leaderboard | — | 63% |
| GDPval | — | 30.8% |
| τ²-bench Banking | 26% | — |
| OSWorld | — | 23% |
| PostTrainBench | 22% | — |
| BALROG | 57% | — |
| ExploitBench | 26.1% | — |
| GBAEval | 0.8% | — |
| GDP.pdf | 17% | — |
| Vending-Bench 2 | 3,774 | — |

## Reasoning

- Gemini 3.1 Pro Preview: 71.7 (#12)
- o3: 32.0 (#78)

| Benchmark | Gemini 3.1 Pro Preview | o3 |
|---|---|---|
| ARC-AGI-2 | 77.1% | 6.5% |
| SimpleBench | 79.6% | 53.1% |
| ARC-AGI-1 | 98% | 60.8% |
| CritPt | 17.7% | 1.4% |
| Chess Puzzles | 55% | 38% |
| EnigmaEval | 36.8% | 13.1% |
| LMArena Hard Prompts | 1485 | 1402 |
| Mystery Game Puzzles | 34% | 29% |
| DTBench | 97.1% | 84.8% |
| LMCA | 53.8% | 39.7% |
| Epoch Capabilities Index | 154.77 | 146.86 |
| ForecastBench | 59 | 62.5 |
| Kagi LLM Benchmark | — | 67.6% |
| NYT Connections (extended) | 97.4% | — |
| Thematic Generalization | 79.4% | — |
| EBR-Bench | 14.3% | — |

## Math

- Gemini 3.1 Pro Preview: 62.1 (#34)
- o3: 50.2 (#58)

| Benchmark | Gemini 3.1 Pro Preview | o3 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 59.6% | 33.3% |
| OTIS Mock AIME 2024-2025 | 95.6% | 84.4% |
| LMArena Math | 1485 | 1426 |
| FrontierMath (Feb 2025 set) | 36.9% | 18.7% |
| FrontierMath Tier 4 (v1) | 16.7% | 2.1% |
| FrontierMath Tier 4 | 26.8% | — |
| MathArena Final-Answer Competitions | 86.5% | — |
| ProofBench | 26% | — |
| Omni-MATH | — | 71.4% |
| MATH Level 5 | — | 97.8% |

## Knowledge

- Gemini 3.1 Pro Preview: 71.8 (#3)
- o3: 54.6 (#52)

| Benchmark | Gemini 3.1 Pro Preview | o3 |
|---|---|---|
| GPQA Diamond | 94.4% | 81.8% |
| Humanity's Last Exam | 46.4% | 20.3% |
| SimpleQA Verified | 73.5% | 49.4% |
| LMArena Expert | 1485 | 1402 |
| MMLU-Pro | — | 85.9% |
| Confabulations | — | 14.4% |
| Vectara Hallucination Rate | 10.4% | — |
| GPQA (HELM) | — | 75.3% |

## Multimodal

- Gemini 3.1 Pro Preview: 37.9 (#69)
- o3: 41.4 (#36)

| Benchmark | Gemini 3.1 Pro Preview | o3 |
|---|---|---|
| LMArena Vision | 1296 | 1214 |
| GeoBench | — | 74% |
| VPCT | — | 52% |
| Blueprint-Bench 2 | 26.5% | — |
| Furniture Assembly | 26.7% | — |
| LMArena Document | 1444 | — |

## Multilingual

- Gemini 3.1 Pro Preview: 57.0 (#12)
- o3: 51.7 (#105)

| Benchmark | Gemini 3.1 Pro Preview | o3 |
|---|---|---|
| LMArena Non-English | 1477 | 1401 |
| LMArena Chinese | 1529 | 1437 |
| LMArena French | 1487 | 1430 |
| LMArena German | 1491 | 1420 |
| LMArena Japanese | 1493 | 1403 |
| LMArena Korean | 1455 | 1370 |
| LMArena Russian | 1498 | 1406 |
| LMArena Spanish | 1479 | 1395 |

## Instruction Following

- Gemini 3.1 Pro Preview: 77.0 (#32)
- o3: 72.8 (#127)

| Benchmark | Gemini 3.1 Pro Preview | o3 |
|---|---|---|
| LMArena Instruction Following | 1466 | 1368 |
| IFEval | — | 86.9% |

## Long Context

- Gemini 3.1 Pro Preview: 47.4 (#18)
- o3: 53.3 (#6)

| Benchmark | Gemini 3.1 Pro Preview | o3 |
|---|---|---|
| CL-bench | 20.8% | 17.8% |
| LMArena Longer Query | 1483 | 1372 |
| Fiction.LiveBench | — | 88.9% |
| CL-bench Life | 16.9% | — |

## Writing & Preference

- Gemini 3.1 Pro Preview: 66.1 (#37)
- o3: 63.5 (#64)

| Benchmark | Gemini 3.1 Pro Preview | o3 |
|---|---|---|
| LMArena Text | 1481 | 1410 |
| LMArena Creative Writing | 1482 | 1359 |
| EQ-Bench Creative Writing | 1491 | 1676 |
| LMArena Multi-Turn | 1488 | 1405 |
| Short-Story Creative Writing | — | 83.9% |
| WildBench | — | 86.1% |
| EQ-Bench 4 | 1142 | — |

## FAQ

### Is Gemini 3.1 Pro Preview better than o3?

Gemini 3.1 Pro Preview is the stronger model overall, scoring 56.7 to 47.5 on the Noometry Index.

### Which is cheaper, Gemini 3.1 Pro Preview or o3?

o3 is cheaper. It lists at $2 per million input tokens and $8 per million output tokens; Gemini 3.1 Pro Preview lists at $2 and $12.

### Is Gemini 3.1 Pro Preview or o3 better for coding?

o3 scores higher on coding benchmarks: 46.8 versus 42.5 in the Noometry coding category.

### Which has the bigger context window?

Gemini 3.1 Pro Preview does, with 1.05M tokens against 200K.

### How many benchmarks do Gemini 3.1 Pro Preview and o3 share?

45 benchmarks have published results for both models. Gemini 3.1 Pro Preview has 71 scored results on Noometry and o3 has 63.