# Gemma 3 12B vs GPT-5 Nano

> GPT-5 Nano is the stronger model overall, scoring 33.5 to 32.1 on the Noometry Index. Gemma 3 12B costs 1.8× less per token, which makes it the better buy when GPT-5 Nano's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/gemma-3-12b-vs-gpt-5-nano
- Last updated: 2026-10-12
- Shared benchmarks: 21

## Summary

- They share 21 benchmarks with published results for both. Gemma 3 12B scores higher in 3 categories and GPT-5 Nano in 6 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where GPT-5 Nano leads 35.9 to 26.5.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 16.7% for Gemma 3 12B and 81.1% for GPT-5 Nano.
- Gemma 3 12B is cheaper at $0.05 / $0.15 per million input/output tokens, against $0.05 / $0.40 for GPT-5 Nano.
- GPT-5 Nano accepts more context: 400K tokens versus 131K.
- Gemma 3 12B has downloadable open weights; the other is API-only.

## Snapshot

| | Gemma 3 12B | GPT-5 Nano |
|---|---|---|
| Provider | Google | OpenAI |
| Noometry Index | 32.1 | 33.5 |
| Rank | 262 | 241 |
| Context | 131K | 400K |
| Input $/M | $0.05 | $0.05 |
| Output $/M | $0.15 | $0.40 |
| Weights | Open | Proprietary |

## Coding

- Gemma 3 12B: 31.7 (#280)
- GPT-5 Nano: 33.6 (#254)

| Benchmark | Gemma 3 12B | GPT-5 Nano |
|---|---|---|
| LMArena Coding | 1281 | 1351 |
| SWE-bench Verified (bash only) | — | 34.8% |
| SciCode | 17.4% | — |
| WeirdML | — | 38.1% |
| ALE-Bench | — | 718.67 |

## Agentic & Tool Use

- Gemma 3 12B: 25.5 (#108)
- GPT-5 Nano: 25.8 (#106)

| Benchmark | Gemma 3 12B | GPT-5 Nano |
|---|---|---|
| Berkeley Function Calling Leaderboard | 30.4% | 51.5% |
| Terminal-Bench | — | 21.8% |

## Reasoning

- Gemma 3 12B: 15.7 (#313)
- GPT-5 Nano: 16.3 (#306)

| Benchmark | Gemma 3 12B | GPT-5 Nano |
|---|---|---|
| Chess Puzzles | 0% | 27% |
| LMArena Hard Prompts | 1309 | 1328 |
| DTBench | 48.8% | 62.7% |
| LMCA | 4.5% | 7.9% |
| Epoch Capabilities Index | 123.5 | 139.38 |
| ARC-AGI-2 | — | 2.6% |
| Kagi LLM Benchmark | — | 62.2% |
| ARC-AGI-1 | — | 20.7% |
| CritPt | 0% | — |
| Mystery Game Puzzles | — | 9% |
| ForecastBench | — | 59.1 |

## Math

- Gemma 3 12B: 22.3 (#279)
- GPT-5 Nano: 29.4 (#241)

| Benchmark | Gemma 3 12B | GPT-5 Nano |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 16.7% | 81.1% |
| LMArena Math | 1307 | 1317 |
| FrontierMath (Tiers 1-3) | — | 20% |
| FrontierMath Tier 4 | — | 2.4% |
| ProofBench | — | 12% |
| Omni-MATH | — | 54.6% |
| MATH Level 5 | — | 95.2% |
| FrontierMath (Feb 2025 set) | — | 8.3% |
| FrontierMath Tier 4 (v1) | — | 2.1% |

## Knowledge

- Gemma 3 12B: 26.5 (#257)
- GPT-5 Nano: 35.9 (#178)

| Benchmark | Gemma 3 12B | GPT-5 Nano |
|---|---|---|
| GPQA Diamond | 39.5% | 69.4% |
| Vectara Hallucination Rate | 4.4% | 10.5% |
| LMArena Expert | 1248 | 1321 |
| SimpleQA Verified | — | 11.7% |
| MMLU-Pro | — | 77.8% |
| GPQA (HELM) | — | 67.9% |

## Multimodal

- Gemma 3 12B: —
- GPT-5 Nano: 31.3 (#108)

| Benchmark | Gemma 3 12B | GPT-5 Nano |
|---|---|---|
| LMArena Vision | — | 1159 |
| VPCT | — | 37.2% |
| MindCube | 46.7% | — |

## Multilingual

- Gemma 3 12B: 45.7 (#165)
- GPT-5 Nano: 45.3 (#172)

| Benchmark | Gemma 3 12B | GPT-5 Nano |
|---|---|---|
| LMArena Non-English | 1318 | 1313 |
| LMArena German | 1370 | 1327 |
| LMArena Russian | 1335 | 1296 |
| LMArena Chinese | — | 1356 |
| LMArena Japanese | — | 1226 |
| LMArena Korean | — | 1269 |
| LMArena Spanish | — | 1360 |

## Instruction Following

- Gemma 3 12B: 68.6 (#186)
- GPT-5 Nano: 75.0 (#79)

| Benchmark | Gemma 3 12B | GPT-5 Nano |
|---|---|---|
| LMArena Instruction Following | 1299 | 1306 |
| IFEval | — | 93.2% |

## Long Context

- Gemma 3 12B: 40.0 (#162)
- GPT-5 Nano: 31.3 (#281)

| Benchmark | Gemma 3 12B | GPT-5 Nano |
|---|---|---|
| LMArena Longer Query | 1317 | 1312 |
| Fiction.LiveBench | — | 44.4% |

## Writing & Preference

- Gemma 3 12B: 47.5 (#209)
- GPT-5 Nano: 39.1 (#249)

| Benchmark | Gemma 3 12B | GPT-5 Nano |
|---|---|---|
| LMArena Text | 1334 | 1320 |
| LMArena Creative Writing | 1331 | 1249 |
| EQ-Bench Creative Writing | 1126 | 705 |
| LMArena Multi-Turn | 1334 | 1311 |
| WildBench | — | 80.6% |

## FAQ

### Is Gemma 3 12B better than GPT-5 Nano?

GPT-5 Nano is the stronger model overall, scoring 33.5 to 32.1 on the Noometry Index. Gemma 3 12B costs 1.8× less per token, which makes it the better buy when GPT-5 Nano's lead doesn't matter for your workload.

### Which is cheaper, Gemma 3 12B or GPT-5 Nano?

Gemma 3 12B is cheaper. It lists at $0.05 per million input tokens and $0.15 per million output tokens; GPT-5 Nano lists at $0.05 and $0.40.

### Is Gemma 3 12B or GPT-5 Nano better for coding?

GPT-5 Nano scores higher on coding benchmarks: 33.6 versus 31.7 in the Noometry coding category.

### Which has the bigger context window?

GPT-5 Nano does, with 400K tokens against 131K.

### How many benchmarks do Gemma 3 12B and GPT-5 Nano share?

21 benchmarks have published results for both models. Gemma 3 12B has 24 scored results on Noometry and GPT-5 Nano has 49.