# Gemini 3.1 Pro Preview vs GPT-6 Astra

> GPT-6 Astra is the stronger model overall, scoring 70.8 to 56.7 on the Noometry Index. Gemini 3.1 Pro Preview costs 4.4× less per token, which makes it the better buy when GPT-6 Astra's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/gemini-3-1-pro-preview-vs-gpt-6-astra
- Last updated: 2026-10-10
- Shared benchmarks: 51

## Summary

- They share 51 benchmarks with published results for both. Gemini 3.1 Pro Preview scores higher in 3 categories and GPT-6 Astra in 7 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-6 Astra leads 93.5 to 62.1.
- The biggest single-benchmark swing is ProofBench: 26% for Gemini 3.1 Pro Preview and 99% for GPT-6 Astra.
- Gemini 3.1 Pro Preview is cheaper at $2 / $12 per million input/output tokens, against $10 / $50 for GPT-6 Astra.
- GPT-6 Astra accepts more context: 1.05M tokens versus 1.05M.

## Snapshot

| | Gemini 3.1 Pro Preview | GPT-6 Astra |
|---|---|---|
| Provider | Google | OpenAI |
| Noometry Index | 56.7 | 70.8 |
| Rank | 23 | 1 |
| Context | 1.05M | 1.05M |
| Input $/M | $2 | $10 |
| Output $/M | $12 | $50 |
| Weights | Proprietary | Proprietary |

## Coding

- Gemini 3.1 Pro Preview: 42.5 (#99)
- GPT-6 Astra: 73.7 (#2)

| Benchmark | Gemini 3.1 Pro Preview | GPT-6 Astra |
|---|---|---|
| DeepSWE | 11.7% | 74.1% |
| LMArena WebDev | 1447 | 1786 |
| SciCode | 58.9% | 56.5% |
| GSO | 22.6% | 79.4% |
| WeirdML | 72.1% | 93.6% |
| LMArena Coding | 1484 | 1487 |
| MirrorCode | 8.9% | 46.7% |
| ALE-Bench | 1,161 | 2,951 |
| SWE-bench Verified | 75.6% | — |
| FrontierCode | — | 53.3% |
| FrontierSWE | — | 65.5% |
| AlgoTune | 2.02 | — |

## Agentic & Tool Use

- Gemini 3.1 Pro Preview: 37.7 (#34)
- GPT-6 Astra: 52.9 (#3)

| Benchmark | Gemini 3.1 Pro Preview | GPT-6 Astra |
|---|---|---|
| APEX-Agents | 35.3% | 64.7% |
| BALROG | 57% | 68.3% |
| GDP.pdf | 17% | 34.2% |
| Vending-Bench 2 | 3,774 | 15,515 |
| Terminal-Bench | 80.2% | — |
| Remote Labor Index | — | 20.8% |
| τ²-bench Banking | 26% | — |
| DeepResearch Bench | 47.8% | — |
| PostTrainBench | 22% | — |
| ExploitBench | 26.1% | — |
| GBAEval | 0.8% | — |
| LMArena Search | 1211 | — |
| METR Time Horizons | 77% | — |

## Reasoning

- Gemini 3.1 Pro Preview: 71.7 (#12)
- GPT-6 Astra: 85.1 (#1)

| Benchmark | Gemini 3.1 Pro Preview | GPT-6 Astra |
|---|---|---|
| ARC-AGI-2 | 77.1% | 95% |
| NYT Connections (extended) | 97.4% | 98.1% |
| ARC-AGI-1 | 98% | 98.5% |
| CritPt | 17.7% | 31.7% |
| Chess Puzzles | 55% | 72% |
| EBR-Bench | 14.3% | 76.2% |
| LMArena Hard Prompts | 1485 | 1462 |
| Mystery Game Puzzles | 34% | 84% |
| DTBench | 97.1% | 97.3% |
| LMCA | 53.8% | 64.4% |
| Epoch Capabilities Index | 154.77 | 166.45 |
| SimpleBench | 79.6% | — |
| EnigmaEval | 36.8% | — |
| Thematic Generalization | 79.4% | — |
| Bench to the Future 3 | — | 0.14 |
| ForecastBench | 59 | — |

## Math

- Gemini 3.1 Pro Preview: 62.1 (#34)
- GPT-6 Astra: 93.5 (#2)

| Benchmark | Gemini 3.1 Pro Preview | GPT-6 Astra |
|---|---|---|
| FrontierMath (Tiers 1-3) | 59.6% | 93.7% |
| FrontierMath Tier 4 | 26.8% | 97.6% |
| OTIS Mock AIME 2024-2025 | 95.6% | 100% |
| ProofBench | 26% | 99% |
| LMArena Math | 1485 | 1465 |
| MathArena Final-Answer Competitions | 86.5% | — |
| FrontierMath (Feb 2025 set) | 36.9% | — |
| FrontierMath Erdős | — | 2.9% |
| FrontierMath Tier 4 (v1) | 16.7% | — |

## Knowledge

- Gemini 3.1 Pro Preview: 71.8 (#3)
- GPT-6 Astra: 75.3 (#1)

| Benchmark | Gemini 3.1 Pro Preview | GPT-6 Astra |
|---|---|---|
| GPQA Diamond | 94.4% | 95.8% |
| Humanity's Last Exam | 46.4% | 54.8% |
| SimpleQA Verified | 73.5% | 75.6% |
| Vectara Hallucination Rate | 10.4% | 8.7% |
| LMArena Expert | 1485 | 1483 |

## Multimodal

- Gemini 3.1 Pro Preview: 37.9 (#69)
- GPT-6 Astra: 55.0 (#3)

| Benchmark | Gemini 3.1 Pro Preview | GPT-6 Astra |
|---|---|---|
| LMArena Vision | 1296 | 1281 |
| Blueprint-Bench 2 | 26.5% | 49.7% |
| Furniture Assembly | 26.7% | 80% |
| LMArena Document | 1444 | 1468 |

## Multilingual

- Gemini 3.1 Pro Preview: 57.0 (#12)
- GPT-6 Astra: 53.7 (#61)

| Benchmark | Gemini 3.1 Pro Preview | GPT-6 Astra |
|---|---|---|
| LMArena Non-English | 1477 | 1430 |
| LMArena Chinese | 1529 | 1484 |
| LMArena French | 1487 | 1456 |
| LMArena German | 1491 | 1440 |
| LMArena Japanese | 1493 | 1379 |
| LMArena Korean | 1455 | 1426 |
| LMArena Russian | 1498 | 1436 |
| LMArena Spanish | 1479 | 1407 |

## Instruction Following

- Gemini 3.1 Pro Preview: 77.0 (#32)
- GPT-6 Astra: 76.3 (#44)

| Benchmark | Gemini 3.1 Pro Preview | GPT-6 Astra |
|---|---|---|
| LMArena Instruction Following | 1466 | 1450 |

## Long Context

- Gemini 3.1 Pro Preview: 47.4 (#18)
- GPT-6 Astra: 44.5 (#62)

| Benchmark | Gemini 3.1 Pro Preview | GPT-6 Astra |
|---|---|---|
| LMArena Longer Query | 1483 | 1456 |
| CL-bench | 20.8% | — |
| CL-bench Life | 16.9% | — |

## Writing & Preference

- Gemini 3.1 Pro Preview: 66.1 (#37)
- GPT-6 Astra: 75.3 (#7)

| Benchmark | Gemini 3.1 Pro Preview | GPT-6 Astra |
|---|---|---|
| LMArena Text | 1481 | 1441 |
| LMArena Creative Writing | 1482 | 1418 |
| EQ-Bench Creative Writing | 1491 | 2173 |
| LMArena Multi-Turn | 1488 | 1448 |
| EQ-Bench 4 | 1142 | — |

## FAQ

### Is Gemini 3.1 Pro Preview better than GPT-6 Astra?

GPT-6 Astra is the stronger model overall, scoring 70.8 to 56.7 on the Noometry Index. Gemini 3.1 Pro Preview costs 4.4× less per token, which makes it the better buy when GPT-6 Astra's lead doesn't matter for your workload.

### Which is cheaper, Gemini 3.1 Pro Preview or GPT-6 Astra?

Gemini 3.1 Pro Preview is cheaper. It lists at $2 per million input tokens and $12 per million output tokens; GPT-6 Astra lists at $10 and $50.

### Is Gemini 3.1 Pro Preview or GPT-6 Astra better for coding?

GPT-6 Astra scores higher on coding benchmarks: 73.7 versus 42.5 in the Noometry coding category.

### Which has the bigger context window?

GPT-6 Astra does, with 1.05M tokens against 1.05M.

### How many benchmarks do Gemini 3.1 Pro Preview and GPT-6 Astra share?

51 benchmarks have published results for both models. Gemini 3.1 Pro Preview has 71 scored results on Noometry and GPT-6 Astra has 56.