# Gemini 3.1 Pro Preview vs Gemini 3.7 Flash

> Gemini 3.7 Flash is the stronger model overall, scoring 59.8 to 56.7 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gemini-3-1-pro-preview-vs-gemini-3-7-flash
- Last updated: 2026-10-10
- Shared benchmarks: 41

## Summary

- They share 41 benchmarks with published results for both. Gemini 3.1 Pro Preview scores higher in 4 categories and Gemini 3.7 Flash in 6 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in coding, where Gemini 3.7 Flash leads 56.2 to 42.5.
- The biggest single-benchmark swing is DeepSWE: 11.7% for Gemini 3.1 Pro Preview and 65.5% for Gemini 3.7 Flash.
- Gemini 3.7 Flash is cheaper at $0.75 / $3.75 per million input/output tokens, against $2 / $12 for Gemini 3.1 Pro Preview.

## Snapshot

| | Gemini 3.1 Pro Preview | Gemini 3.7 Flash |
|---|---|---|
| Provider | Google | Google |
| Noometry Index | 56.7 | 59.8 |
| Rank | 23 | 14 |
| Context | 1.05M | 1.05M |
| Input $/M | $2 | $0.75 |
| Output $/M | $12 | $3.75 |
| Weights | Proprietary | Proprietary |

## Coding

- Gemini 3.1 Pro Preview: 42.5 (#99)
- Gemini 3.7 Flash: 56.2 (#22)

| Benchmark | Gemini 3.1 Pro Preview | Gemini 3.7 Flash |
|---|---|---|
| DeepSWE | 11.7% | 65.5% |
| LMArena WebDev | 1447 | 1592 |
| SciCode | 58.9% | 59.8% |
| LMArena Coding | 1484 | 1497 |
| ALE-Bench | 1,161 | 904.3 |
| SWE-bench Verified | 75.6% | — |
| FrontierCode | — | 43.6% |
| FrontierSWE | — | 20.3% |
| GSO | 22.6% | — |
| WeirdML | 72.1% | — |
| MirrorCode | 8.9% | — |
| AlgoTune | 2.02 | — |

## Agentic & Tool Use

- Gemini 3.1 Pro Preview: 37.7 (#34)
- Gemini 3.7 Flash: 42.1 (#19)

| Benchmark | Gemini 3.1 Pro Preview | Gemini 3.7 Flash |
|---|---|---|
| APEX-Agents | 35.3% | 67.8% |
| GDP.pdf | 17% | 23.8% |
| Terminal-Bench | 80.2% | — |
| Remote Labor Index | — | 5% |
| τ²-bench Banking | 26% | — |
| DeepResearch Bench | 47.8% | — |
| PostTrainBench | 22% | — |
| BALROG | 57% | — |
| ExploitBench | 26.1% | — |
| GBAEval | 0.8% | — |
| LMArena Search | 1211 | — |
| METR Time Horizons | 77% | — |
| Vending-Bench 2 | 3,774 | — |

## Reasoning

- Gemini 3.1 Pro Preview: 71.7 (#12)
- Gemini 3.7 Flash: 70.0 (#15)

| Benchmark | Gemini 3.1 Pro Preview | Gemini 3.7 Flash |
|---|---|---|
| ARC-AGI-2 | 77.1% | 84.6% |
| NYT Connections (extended) | 97.4% | 94% |
| ARC-AGI-1 | 98% | 95.5% |
| CritPt | 17.7% | 14.3% |
| Chess Puzzles | 55% | 47% |
| LMArena Hard Prompts | 1485 | 1494 |
| Mystery Game Puzzles | 34% | 37% |
| DTBench | 97.1% | 96.8% |
| LMCA | 53.8% | 50.4% |
| Epoch Capabilities Index | 154.77 | 157.27 |
| SimpleBench | 79.6% | — |
| EnigmaEval | 36.8% | — |
| Thematic Generalization | 79.4% | — |
| EBR-Bench | 14.3% | — |
| ForecastBench | 59 | — |

## Math

- Gemini 3.1 Pro Preview: 62.1 (#34)
- Gemini 3.7 Flash: 69.6 (#23)

| Benchmark | Gemini 3.1 Pro Preview | Gemini 3.7 Flash |
|---|---|---|
| FrontierMath (Tiers 1-3) | 59.6% | 71.6% |
| FrontierMath Tier 4 | 26.8% | 36.6% |
| OTIS Mock AIME 2024-2025 | 95.6% | 97.2% |
| ProofBench | 26% | 58% |
| LMArena Math | 1485 | 1507 |
| MathArena Final-Answer Competitions | 86.5% | — |
| FrontierMath (Feb 2025 set) | 36.9% | — |
| FrontierMath Tier 4 (v1) | 16.7% | — |

## Knowledge

- Gemini 3.1 Pro Preview: 71.8 (#3)
- Gemini 3.7 Flash: 69.7 (#5)

| Benchmark | Gemini 3.1 Pro Preview | Gemini 3.7 Flash |
|---|---|---|
| GPQA Diamond | 94.4% | 94.8% |
| SimpleQA Verified | 73.5% | 69.2% |
| LMArena Expert | 1485 | 1508 |
| Humanity's Last Exam | 46.4% | — |
| Vectara Hallucination Rate | 10.4% | — |

## Multimodal

- Gemini 3.1 Pro Preview: 37.9 (#69)
- Gemini 3.7 Flash: 37.3 (#73)

| Benchmark | Gemini 3.1 Pro Preview | Gemini 3.7 Flash |
|---|---|---|
| LMArena Vision | 1296 | 1316 |
| Furniture Assembly | 26.7% | 26.7% |
| Blueprint-Bench 2 | 26.5% | — |
| LMArena Document | 1444 | — |

## Multilingual

- Gemini 3.1 Pro Preview: 57.0 (#12)
- Gemini 3.7 Flash: 57.6 (#7)

| Benchmark | Gemini 3.1 Pro Preview | Gemini 3.7 Flash |
|---|---|---|
| LMArena Non-English | 1477 | 1484 |
| LMArena Chinese | 1529 | 1548 |
| LMArena French | 1487 | 1505 |
| LMArena German | 1491 | 1498 |
| LMArena Japanese | 1493 | 1512 |
| LMArena Korean | 1455 | 1483 |
| LMArena Russian | 1498 | 1516 |
| LMArena Spanish | 1479 | 1503 |

## Instruction Following

- Gemini 3.1 Pro Preview: 77.0 (#32)
- Gemini 3.7 Flash: 77.7 (#15)

| Benchmark | Gemini 3.1 Pro Preview | Gemini 3.7 Flash |
|---|---|---|
| LMArena Instruction Following | 1466 | 1483 |

## Long Context

- Gemini 3.1 Pro Preview: 47.4 (#18)
- Gemini 3.7 Flash: 45.7 (#30)

| Benchmark | Gemini 3.1 Pro Preview | Gemini 3.7 Flash |
|---|---|---|
| LMArena Longer Query | 1483 | 1492 |
| CL-bench | 20.8% | — |
| CL-bench Life | 16.9% | — |

## Writing & Preference

- Gemini 3.1 Pro Preview: 66.1 (#37)
- Gemini 3.7 Flash: 71.2 (#20)

| Benchmark | Gemini 3.1 Pro Preview | Gemini 3.7 Flash |
|---|---|---|
| LMArena Text | 1481 | 1486 |
| LMArena Creative Writing | 1482 | 1490 |
| EQ-Bench Creative Writing | 1491 | 1723 |
| LMArena Multi-Turn | 1488 | 1489 |
| EQ-Bench 4 | 1142 | — |

## FAQ

### Is Gemini 3.1 Pro Preview better than Gemini 3.7 Flash?

Gemini 3.7 Flash is the stronger model overall, scoring 59.8 to 56.7 on the Noometry Index.

### Which is cheaper, Gemini 3.1 Pro Preview or Gemini 3.7 Flash?

Gemini 3.7 Flash is cheaper. It lists at $0.75 per million input tokens and $3.75 per million output tokens; Gemini 3.1 Pro Preview lists at $2 and $12.

### Is Gemini 3.1 Pro Preview or Gemini 3.7 Flash better for coding?

Gemini 3.7 Flash scores higher on coding benchmarks: 56.2 versus 42.5 in the Noometry coding category.

### Which has the bigger context window?

Both accept 1.05M tokens.

### How many benchmarks do Gemini 3.1 Pro Preview and Gemini 3.7 Flash share?

41 benchmarks have published results for both models. Gemini 3.1 Pro Preview has 71 scored results on Noometry and Gemini 3.7 Flash has 44.