Model comparison
Gemma 3 27B vs GPT-5.4 mini
GPT-5.4 mini is the stronger model overall, scoring 45.0 to 30.8 on the Noometry Index. Gemma 3 27B costs 17× less per token, which makes it the better buy when GPT-5.4 mini's lead doesn't matter for your workload.
Last verified . 29 shared benchmarks.
GoogleSummary
- They share 29 benchmarks with published results for both. Gemma 3 27B scores higher in 0 categories and GPT-5.4 mini in 10 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where GPT-5.4 mini leads 51.5 to 25.5.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 22.5% for Gemma 3 27B and 88.9% for GPT-5.4 mini.
- Gemma 3 27B is cheaper at $0.08 / $0.16 per million input/output tokens, against $0.75 / $4.50 for GPT-5.4 mini.
- GPT-5.4 mini 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-5.4 mini | |
|---|---|---|
| Provider | OpenAI | |
| Noometry Index | 30.8 | 45.0 |
| Released | 2025-03-11 | 2026-03-17 |
| Weights | Open | Proprietary |
| Context window | 131K | 400K |
| Max output | 8K | 128K |
| Input $ / M tokens | $0.08 | $0.75 |
| Output $ / M tokens | $0.16 | $4.50 |
| Results tracked | 43 | 46 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.4 mini leads
Gemma 3 27B: 22.5 (#334), GPT-5.4 mini: 45.2 (#72)
| Benchmark | Gemma 3 27B | GPT-5.4 mini |
|---|---|---|
| SciCode | 21.2% | 49.9% |
| LMArena Coding | 1322 | 1438 |
| FrontierCode | — | 27% |
| Aider Polyglot | 4.9% | — |
| LMArena WebDev | — | 1397 |
| WeirdML | — | 60.3% |
| LiveBench Coding | 39.9% | — |
| ALE-Bench | — | 1,189 |
Agentic & Tool Use GPT-5.4 mini leads
Gemma 3 27B: 25.1 (#110), GPT-5.4 mini: 29.9 (#81)
| Benchmark | Gemma 3 27B | GPT-5.4 mini |
|---|---|---|
| Berkeley Function Calling Leaderboard | 29.5% | — |
| DeepResearch Bench | — | 36.3% |
Reasoning GPT-5.4 mini leads
Gemma 3 27B: 16.7 (#301), GPT-5.4 mini: 30.4 (#85)
| Benchmark | Gemma 3 27B | GPT-5.4 mini |
|---|---|---|
| Kagi LLM Benchmark | 40.4% | 37.9% |
| CritPt | 0% | 10% |
| Chess Puzzles | 0% | 24% |
| LMArena Hard Prompts | 1340 | 1424 |
| DTBench | 52.5% | 80% |
| LMCA | 12.3% | 40.8% |
| Epoch Capabilities Index | 130.04 | 148.84 |
| ARC-AGI-2 | — | 18.9% |
| NYT Connections (extended) | — | 61.8% |
| ARC-AGI-1 | — | 63.7% |
| Thematic Generalization | — | 61.7% |
| LiveBench Reasoning | 43.8% | — |
| Mystery Game Puzzles | — | 11% |
| LiveBench Data Analysis | 51.5% | — |
| ForecastBench | — | 57 |
| LiveBench | 50% | — |
Math GPT-5.4 mini leads
Gemma 3 27B: 25.9 (#265), GPT-5.4 mini: 45.5 (#75)
| Benchmark | Gemma 3 27B | GPT-5.4 mini |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 22.5% | 88.9% |
| LMArena Math | 1312 | 1419 |
| FrontierMath (Tiers 1-3) | — | 51.2% |
| FrontierMath Tier 4 | — | 9.8% |
| ProofBench | — | 21% |
| LiveBench Math | 55.4% | — |
| MATH Level 5 | 74% | — |
| FrontierMath (Feb 2025 set) | — | 28.3% |
| FrontierMath Tier 4 (v1) | — | 2.1% |
Knowledge GPT-5.4 mini leads
Gemma 3 27B: 25.5 (#261), GPT-5.4 mini: 51.5 (#67)
| Benchmark | Gemma 3 27B | GPT-5.4 mini |
|---|---|---|
| GPQA Diamond | 47.7% | 86.9% |
| Vectara Hallucination Rate | 7.4% | 5.5% |
| LMArena Expert | 1304 | 1435 |
| SimpleQA Verified | — | 29.4% |
| Confabulations | 40.3% | — |
Multimodal GPT-5.4 mini leads
Gemma 3 27B: 32.6 (#100), GPT-5.4 mini: 39.7 (#56)
| Benchmark | Gemma 3 27B | GPT-5.4 mini |
|---|---|---|
| LMArena Vision | 1164 | 1245 |
| GeoBench | 52% | — |
Multilingual GPT-5.4 mini leads
Gemma 3 27B: 46.9 (#155), GPT-5.4 mini: 51.9 (#96)
| Benchmark | Gemma 3 27B | GPT-5.4 mini |
|---|---|---|
| LMArena Non-English | 1334 | 1405 |
| LMArena Chinese | 1346 | 1446 |
| LMArena French | 1368 | 1440 |
| LMArena German | 1362 | 1409 |
| LMArena Japanese | 1287 | 1374 |
| LMArena Korean | 1308 | 1368 |
| LMArena Russian | 1349 | 1417 |
| LMArena Spanish | 1349 | 1405 |
Instruction Following GPT-5.4 mini leads
Gemma 3 27B: 70.6 (#160), GPT-5.4 mini: 74.1 (#102)
| Benchmark | Gemma 3 27B | GPT-5.4 mini |
|---|---|---|
| LMArena Instruction Following | 1321 | 1405 |
| LiveBench Instruction Following | 74.9% | — |
Long Context GPT-5.4 mini leads
Gemma 3 27B: 27.6 (#293), GPT-5.4 mini: 43.0 (#112)
| Benchmark | Gemma 3 27B | GPT-5.4 mini |
|---|---|---|
| LMArena Longer Query | 1333 | 1407 |
| Fiction.LiveBench | 33.3% | — |
Writing & Preference GPT-5.4 mini leads
Gemma 3 27B: 52.5 (#168), GPT-5.4 mini: 64.0 (#58)
| Benchmark | Gemma 3 27B | GPT-5.4 mini |
|---|---|---|
| LMArena Text | 1358 | 1412 |
| LMArena Creative Writing | 1346 | 1370 |
| EQ-Bench Creative Writing | 1266 | 1665 |
| LMArena Multi-Turn | 1345 | 1429 |
| Short-Story Creative Writing | 79.9% | — |
| LiveBench Language | 34.6% | — |
Frequently asked questions
Is Gemma 3 27B better than GPT-5.4 mini?
GPT-5.4 mini is the stronger model overall, scoring 45.0 to 30.8 on the Noometry Index. Gemma 3 27B costs 17× less per token, which makes it the better buy when GPT-5.4 mini's lead doesn't matter for your workload.
Which is cheaper, Gemma 3 27B or GPT-5.4 mini?
Gemma 3 27B is cheaper. It lists at $0.08 per million input tokens and $0.16 per million output tokens; GPT-5.4 mini lists at $0.75 and $4.50.
Is Gemma 3 27B or GPT-5.4 mini better for coding?
GPT-5.4 mini scores higher on coding benchmarks: 45.2 versus 22.5 in the Noometry coding category.
Which has the bigger context window?
GPT-5.4 mini does, with 400K tokens against 131K.
How many benchmarks do Gemma 3 27B and GPT-5.4 mini share?
29 benchmarks have published results for both models. Gemma 3 27B has 43 scored results on Noometry and GPT-5.4 mini has 46.