Model comparison
Gemma 2 27B vs GPT-5.4 nano
GPT-5.4 nano is the stronger model overall, scoring 41.9 to 29.4 on the Noometry Index.
Last verified . 22 shared benchmarks.
GoogleSummary
- They share 22 benchmarks with published results for both. Gemma 2 27B scores higher in 0 categories and GPT-5.4 nano in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.4 nano leads 40.9 to 10.7.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 1.4% for Gemma 2 27B and 87.8% for GPT-5.4 nano.
- GPT-5.4 nano is cheaper at $0.20 / $1.25 per million input/output tokens, against $0.65 / $0.65 for Gemma 2 27B.
- GPT-5.4 nano accepts more context: 400K tokens versus 8K.
- Gemma 2 27B has downloadable open weights; the other is API-only.
Side by side
| Gemma 2 27B | GPT-5.4 nano | |
|---|---|---|
| Provider | OpenAI | |
| Noometry Index | 29.4 | 41.9 |
| Released | 2024-06-24 | 2026-03-17 |
| Weights | Open | Proprietary |
| Context window | 8K | 400K |
| Max output | 2K | 128K |
| Input $ / M tokens | $0.65 | $0.20 |
| Output $ / M tokens | $0.65 | $1.25 |
| Results tracked | 34 | 40 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.4 nano leads
Gemma 2 27B: 34.1 (#246), GPT-5.4 nano: 43.6 (#84)
| Benchmark | Gemma 2 27B | GPT-5.4 nano |
|---|---|---|
| LMArena Coding | 1211 | 1405 |
| SciCode | — | 46.9% |
| WeirdML | — | 49.2% |
| BigCodeBench Instruct | 42.8% | — |
| LiveBench Coding | 36% | — |
| BigCodeBench Complete | 52.5% | — |
| ALE-Bench | — | 1,005 |
Reasoning GPT-5.4 nano leads
Gemma 2 27B: 15.3 (#315), GPT-5.4 nano: 23.7 (#173)
| Benchmark | Gemma 2 27B | GPT-5.4 nano |
|---|---|---|
| LMArena Hard Prompts | 1198 | 1381 |
| DTBench | 48% | 80.3% |
| LMCA | 7.1% | 36.9% |
| Epoch Capabilities Index | 122.08 | 145.81 |
| ARC-AGI-2 | — | 5.7% |
| Kagi LLM Benchmark | — | 39.7% |
| ARC-AGI-1 | — | 51.5% |
| CritPt | — | 9.3% |
| Chess Puzzles | — | 30% |
| LiveBench Reasoning | 28.1% | — |
| Mystery Game Puzzles | — | 9% |
| LiveBench Data Analysis | 47.9% | — |
| ForecastBench | — | 57.3 |
| LiveBench | 38.2% | — |
Math GPT-5.4 nano leads
Gemma 2 27B: 10.7 (#311), GPT-5.4 nano: 40.9 (#88)
| Benchmark | Gemma 2 27B | GPT-5.4 nano |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 1.4% | 87.8% |
| LMArena Math | 1212 | 1406 |
| FrontierMath (Tiers 1-3) | — | 44.9% |
| FrontierMath Tier 4 | — | 12.2% |
| ProofBench | — | 5% |
| LiveBench Math | 26.5% | — |
| MATH Level 5 | 27.9% | — |
| FrontierMath (Feb 2025 set) | — | 25.9% |
| FrontierMath Tier 4 (v1) | — | 6.3% |
Knowledge GPT-5.4 nano leads
Gemma 2 27B: 19.0 (#280), GPT-5.4 nano: 41.9 (#103)
| Benchmark | Gemma 2 27B | GPT-5.4 nano |
|---|---|---|
| GPQA Diamond | 36.5% | 78.5% |
| LMArena Expert | 1172 | 1396 |
| SimpleQA Verified | — | 11.7% |
| Confabulations | 27.1% | — |
| Vectara Hallucination Rate | — | 3.1% |
| MMLU | 75.7% | — |
Multimodal Not comparable
Gemma 2 27B: —, GPT-5.4 nano: 36.7 (#78)
| Benchmark | Gemma 2 27B | GPT-5.4 nano |
|---|---|---|
| LMArena Vision | — | 1196 |
Multilingual GPT-5.4 nano leads
Gemma 2 27B: 38.6 (#226), GPT-5.4 nano: 48.6 (#140)
| Benchmark | Gemma 2 27B | GPT-5.4 nano |
|---|---|---|
| LMArena Non-English | 1217 | 1359 |
| LMArena Chinese | 1221 | 1392 |
| LMArena French | 1247 | 1396 |
| LMArena German | 1209 | 1367 |
| LMArena Japanese | 1175 | 1343 |
| LMArena Korean | 1174 | 1320 |
| LMArena Russian | 1234 | 1363 |
| LMArena Spanish | 1228 | 1371 |
Instruction Following GPT-5.4 nano leads
Gemma 2 27B: 60.5 (#249), GPT-5.4 nano: 71.9 (#144)
| Benchmark | Gemma 2 27B | GPT-5.4 nano |
|---|---|---|
| LMArena Instruction Following | 1206 | 1362 |
| LiveBench Instruction Following | 58.1% | — |
Long Context GPT-5.4 nano leads
Gemma 2 27B: 37.3 (#218), GPT-5.4 nano: 41.6 (#137)
| Benchmark | Gemma 2 27B | GPT-5.4 nano |
|---|---|---|
| LMArena Longer Query | 1231 | 1366 |
Writing & Preference GPT-5.4 nano leads
Gemma 2 27B: 44.2 (#225), GPT-5.4 nano: 55.7 (#142)
| Benchmark | Gemma 2 27B | GPT-5.4 nano |
|---|---|---|
| LMArena Text | 1231 | 1372 |
| LMArena Creative Writing | 1241 | 1314 |
| LMArena Multi-Turn | 1224 | 1382 |
| LiveBench Language | 32.6% | — |
Frequently asked questions
Is Gemma 2 27B better than GPT-5.4 nano?
GPT-5.4 nano is the stronger model overall, scoring 41.9 to 29.4 on the Noometry Index.
Which is cheaper, Gemma 2 27B or GPT-5.4 nano?
GPT-5.4 nano is cheaper. It lists at $0.20 per million input tokens and $1.25 per million output tokens; Gemma 2 27B lists at $0.65 and $0.65.
Is Gemma 2 27B or GPT-5.4 nano better for coding?
GPT-5.4 nano scores higher on coding benchmarks: 43.6 versus 34.1 in the Noometry coding category.
Which has the bigger context window?
GPT-5.4 nano does, with 400K tokens against 8K.
How many benchmarks do Gemma 2 27B and GPT-5.4 nano share?
22 benchmarks have published results for both models. Gemma 2 27B has 34 scored results on Noometry and GPT-5.4 nano has 40.