Model comparison
Gemma 3 27B vs GPT-5 Nano
GPT-5 Nano is the stronger model overall, scoring 33.4 to 30.5 on the Noometry Index.
Last verified . 28 shared benchmarks.
GoogleSummary
- They share 28 benchmarks with published results for both. Gemma 3 27B scores higher in 3 categories and GPT-5 Nano in 7 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Gemma 3 27B leads 52.7 to 39.4.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 22.5% for Gemma 3 27B and 81.1% for GPT-5 Nano.
- Gemma 3 27B is cheaper at $0.08 / $0.16 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 27B has downloadable open weights; the other is API-only.
Side by side
| Gemma 3 27B | GPT-5 Nano | |
|---|---|---|
| Provider | OpenAI | |
| Noometry Index | 30.5 | 33.4 |
| Released | 2025-03-11 | 2025-08-07 |
| Weights | Open | Proprietary |
| Context window | 131K | 400K |
| Max output | 8K | 128K |
| Input $ / M tokens | $0.08 | $0.05 |
| Output $ / M tokens | $0.16 | $0.40 |
| Results tracked | 43 | 48 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5 Nano leads
Gemma 3 27B: 22.1 (#329), GPT-5 Nano: 33.3 (#248)
| Benchmark | Gemma 3 27B | GPT-5 Nano |
|---|---|---|
| LMArena Coding | 1322 | 1351 |
| SWE-bench Verified (bash only) | — | 34.8% |
| Aider Polyglot | 4.9% | — |
| SciCode | 21.2% | — |
| WeirdML | — | 38.1% |
| LiveBench Coding | 39.9% | — |
| ALE-Bench | — | 718.67 |
Agentic & Tool Use Too close to call
Gemma 3 27B: 24.9 (#105), GPT-5 Nano: 25.6 (#99)
| Benchmark | Gemma 3 27B | GPT-5 Nano |
|---|---|---|
| Berkeley Function Calling Leaderboard | 29.5% | 51.5% |
| Terminal-Bench | — | 21.8% |
Reasoning Too close to call
Gemma 3 27B: 16.3 (#297), GPT-5 Nano: 16.4 (#291)
| Benchmark | Gemma 3 27B | GPT-5 Nano |
|---|---|---|
| Kagi LLM Benchmark | 40.4% | 62.2% |
| Chess Puzzles | 0% | 27% |
| LMArena Hard Prompts | 1340 | 1329 |
| DTBench | 52.5% | 62.7% |
| LMCA | 12.3% | 7.9% |
| Epoch Capabilities Index | 130.04 | 139.39 |
| ARC-AGI-2 | — | 2.6% |
| ARC-AGI-1 | — | 20.7% |
| CritPt | 0% | — |
| LiveBench Reasoning | 43.8% | — |
| Mystery Game Puzzles | — | 9% |
| LiveBench Data Analysis | 51.5% | — |
| ForecastBench | — | 59.1 |
| LiveBench | 50% | — |
Math GPT-5 Nano leads
Gemma 3 27B: 24.4 (#261), GPT-5 Nano: 29.1 (#221)
| Benchmark | Gemma 3 27B | GPT-5 Nano |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 22.5% | 81.1% |
| LMArena Math | 1312 | 1318 |
| MATH Level 5 | 74% | 95.2% |
| FrontierMath (Tiers 1-3) | — | 20% |
| FrontierMath Tier 4 | — | 2.4% |
| ProofBench | — | 12% |
| Omni-MATH | — | 54.6% |
| LiveBench Math | 55.4% | — |
| FrontierMath (Feb 2025 set) | — | 8.3% |
| FrontierMath Tier 4 (v1) | — | 2.1% |
Knowledge GPT-5 Nano leads
Gemma 3 27B: 25.4 (#256), GPT-5 Nano: 35.8 (#166)
| Benchmark | Gemma 3 27B | GPT-5 Nano |
|---|---|---|
| GPQA Diamond | 47.7% | 69.4% |
| Vectara Hallucination Rate | 7.4% | 10.5% |
| LMArena Expert | 1303 | 1321 |
| SimpleQA Verified | — | 11.7% |
| MMLU-Pro | — | 77.8% |
| Confabulations | 40.3% | — |
| GPQA (HELM) | — | 67.9% |
Multimodal Gemma 3 27B leads
Gemma 3 27B: 32.0 (#94), GPT-5 Nano: 30.7 (#102)
| Benchmark | Gemma 3 27B | GPT-5 Nano |
|---|---|---|
| LMArena Vision | 1165 | 1159 |
| GeoBench | 52% | — |
| VPCT | — | 37.2% |
Multilingual Gemma 3 27B leads
Gemma 3 27B: 47.0 (#150), GPT-5 Nano: 45.5 (#167)
| Benchmark | Gemma 3 27B | GPT-5 Nano |
|---|---|---|
| LMArena Non-English | 1334 | 1313 |
| LMArena Chinese | 1347 | 1355 |
| LMArena German | 1362 | 1329 |
| LMArena Korean | 1308 | 1269 |
| LMArena Russian | 1349 | 1296 |
| LMArena Spanish | 1348 | 1360 |
| LMArena French | 1368 | — |
| LMArena Japanese | 1287 | — |
Instruction Following GPT-5 Nano leads
Gemma 3 27B: 70.8 (#155), GPT-5 Nano: 75.2 (#72)
| Benchmark | Gemma 3 27B | GPT-5 Nano |
|---|---|---|
| LMArena Instruction Following | 1321 | 1306 |
| LiveBench Instruction Following | 74.9% | — |
| IFEval | — | 93.2% |
Long Context GPT-5 Nano leads
Gemma 3 27B: 27.6 (#288), GPT-5 Nano: 31.3 (#275)
| Benchmark | Gemma 3 27B | GPT-5 Nano |
|---|---|---|
| Fiction.LiveBench | 33.3% | 44.4% |
| LMArena Longer Query | 1333 | 1313 |
Writing & Preference Gemma 3 27B leads
Gemma 3 27B: 52.7 (#161), GPT-5 Nano: 39.4 (#244)
| Benchmark | Gemma 3 27B | GPT-5 Nano |
|---|---|---|
| LMArena Text | 1358 | 1320 |
| LMArena Creative Writing | 1346 | 1250 |
| EQ-Bench Creative Writing | 1266 | 705 |
| LMArena Multi-Turn | 1344 | 1311 |
| Short-Story Creative Writing | 79.9% | — |
| WildBench | — | 80.6% |
| LiveBench Language | 34.6% | — |
Frequently asked questions
Is Gemma 3 27B better than GPT-5 Nano?
GPT-5 Nano is the stronger model overall, scoring 33.4 to 30.5 on the Noometry Index.
Which is cheaper, Gemma 3 27B or GPT-5 Nano?
Gemma 3 27B is cheaper. It lists at $0.08 per million input tokens and $0.16 per million output tokens; GPT-5 Nano lists at $0.05 and $0.40.
Is Gemma 3 27B or GPT-5 Nano better for coding?
GPT-5 Nano scores higher on coding benchmarks: 33.3 versus 22.1 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 27B and GPT-5 Nano share?
28 benchmarks have published results for both models. Gemma 3 27B has 43 scored results on Noometry and GPT-5 Nano has 48.