Model comparison
Gemma 3 27B vs GPT-6 Sol
GPT-6 Sol is the stronger model overall, scoring 61.8 to 30.8 on the Noometry Index. Gemma 3 27B costs 40× less per token, which makes it the better buy when GPT-6 Sol's lead doesn't matter for your workload.
Last verified . 27 shared benchmarks.
GoogleSummary
- They share 27 benchmarks with published results for both. Gemma 3 27B scores higher in 0 categories and GPT-6 Sol in 10 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-6 Sol leads 87.2 to 25.9.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 22.5% for Gemma 3 27B and 100% for GPT-6 Sol.
- Gemma 3 27B is cheaper at $0.08 / $0.16 per million input/output tokens, against $2 / $10 for GPT-6 Sol.
- GPT-6 Sol accepts more context: 1.05M tokens versus 131K.
- Gemma 3 27B has downloadable open weights; the other is API-only.
Side by side
| Gemma 3 27B | GPT-6 Sol | |
|---|---|---|
| Provider | OpenAI | |
| Noometry Index | 30.8 | 61.8 |
| Released | 2025-03-11 | 2026-09-22 |
| Weights | Open | Proprietary |
| Context window | 131K | 1.05M |
| Max output | 8K | 128K |
| Input $ / M tokens | $0.08 | $2 |
| Output $ / M tokens | $0.16 | $10 |
| Results tracked | 43 | 45 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-6 Sol leads
Gemma 3 27B: 22.5 (#334), GPT-6 Sol: 60.1 (#11)
| Benchmark | Gemma 3 27B | GPT-6 Sol |
|---|---|---|
| SciCode | 21.2% | 57.6% |
| LMArena Coding | 1322 | 1447 |
| DeepSWE | — | 68.8% |
| FrontierCode | — | 49.3% |
| Aider Polyglot | 4.9% | — |
| LMArena WebDev | — | 1688 |
| LiveBench Coding | 39.9% | — |
| ALE-Bench | — | 2,462 |
Agentic & Tool Use GPT-6 Sol leads
Gemma 3 27B: 25.1 (#110), GPT-6 Sol: 37.2 (#36)
| Benchmark | Gemma 3 27B | GPT-6 Sol |
|---|---|---|
| APEX-Agents | — | 54.3% |
| Berkeley Function Calling Leaderboard | 29.5% | — |
| GDP.pdf | — | 26.4% |
| Vending-Bench 2 | — | 14,428 |
Reasoning GPT-6 Sol leads
Gemma 3 27B: 16.7 (#301), GPT-6 Sol: 74.0 (#9)
| Benchmark | Gemma 3 27B | GPT-6 Sol |
|---|---|---|
| CritPt | 0% | 30.9% |
| LMArena Hard Prompts | 1340 | 1418 |
| DTBench | 52.5% | 97.3% |
| LMCA | 12.3% | 59.1% |
| Epoch Capabilities Index | 130.04 | 162.72 |
| ARC-AGI-2 | — | 89.6% |
| Kagi LLM Benchmark | 40.4% | — |
| NYT Connections (extended) | — | 90.1% |
| ARC-AGI-1 | — | 95.5% |
| Chess Puzzles | 0% | — |
| EBR-Bench | — | 53.3% |
| LiveBench Reasoning | 43.8% | — |
| Mystery Game Puzzles | — | 56% |
| LiveBench Data Analysis | 51.5% | — |
| LiveBench | 50% | — |
Math GPT-6 Sol leads
Gemma 3 27B: 25.9 (#265), GPT-6 Sol: 87.2 (#7)
| Benchmark | Gemma 3 27B | GPT-6 Sol |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 22.5% | 100% |
| LMArena Math | 1312 | 1402 |
| FrontierMath (Tiers 1-3) | — | 89.8% |
| FrontierMath Tier 4 | — | 90% |
| ProofBench | — | 83% |
| LiveBench Math | 55.4% | — |
| MATH Level 5 | 74% | — |
Knowledge GPT-6 Sol leads
Gemma 3 27B: 25.5 (#261), GPT-6 Sol: 64.8 (#15)
| Benchmark | Gemma 3 27B | GPT-6 Sol |
|---|---|---|
| GPQA Diamond | 47.7% | 94.3% |
| Vectara Hallucination Rate | 7.4% | 6.5% |
| LMArena Expert | 1304 | 1439 |
| SimpleQA Verified | — | 60.7% |
| Confabulations | 40.3% | — |
Multimodal GPT-6 Sol leads
Gemma 3 27B: 32.6 (#100), GPT-6 Sol: 47.6 (#10)
| Benchmark | Gemma 3 27B | GPT-6 Sol |
|---|---|---|
| LMArena Vision | 1164 | 1245 |
| GeoBench | 52% | — |
| Blueprint-Bench 2 | — | 36.9% |
| Furniture Assembly | — | 58.3% |
Multilingual GPT-6 Sol leads
Gemma 3 27B: 46.9 (#155), GPT-6 Sol: 50.5 (#118)
| Benchmark | Gemma 3 27B | GPT-6 Sol |
|---|---|---|
| LMArena Non-English | 1334 | 1385 |
| LMArena Chinese | 1346 | 1405 |
| LMArena French | 1368 | 1410 |
| LMArena German | 1362 | 1390 |
| LMArena Japanese | 1287 | 1385 |
| LMArena Korean | 1308 | 1341 |
| LMArena Russian | 1349 | 1401 |
| LMArena Spanish | 1349 | 1384 |
Instruction Following GPT-6 Sol leads
Gemma 3 27B: 70.6 (#160), GPT-6 Sol: 74.5 (#94)
| Benchmark | Gemma 3 27B | GPT-6 Sol |
|---|---|---|
| LMArena Instruction Following | 1321 | 1412 |
| LiveBench Instruction Following | 74.9% | — |
Long Context GPT-6 Sol leads
Gemma 3 27B: 27.6 (#293), GPT-6 Sol: 43.1 (#108)
| Benchmark | Gemma 3 27B | GPT-6 Sol |
|---|---|---|
| LMArena Longer Query | 1333 | 1411 |
| Fiction.LiveBench | 33.3% | — |
Writing & Preference GPT-6 Sol leads
Gemma 3 27B: 52.5 (#168), GPT-6 Sol: 71.9 (#18)
| Benchmark | Gemma 3 27B | GPT-6 Sol |
|---|---|---|
| LMArena Text | 1358 | 1395 |
| LMArena Creative Writing | 1346 | 1378 |
| EQ-Bench Creative Writing | 1266 | 2125 |
| LMArena Multi-Turn | 1345 | 1412 |
| Short-Story Creative Writing | 79.9% | — |
| LiveBench Language | 34.6% | — |
Frequently asked questions
Is Gemma 3 27B better than GPT-6 Sol?
GPT-6 Sol is the stronger model overall, scoring 61.8 to 30.8 on the Noometry Index. Gemma 3 27B costs 40× less per token, which makes it the better buy when GPT-6 Sol's lead doesn't matter for your workload.
Which is cheaper, Gemma 3 27B or GPT-6 Sol?
Gemma 3 27B is cheaper. It lists at $0.08 per million input tokens and $0.16 per million output tokens; GPT-6 Sol lists at $2 and $10.
Is Gemma 3 27B or GPT-6 Sol better for coding?
GPT-6 Sol scores higher on coding benchmarks: 60.1 versus 22.5 in the Noometry coding category.
Which has the bigger context window?
GPT-6 Sol does, with 1.05M tokens against 131K.
How many benchmarks do Gemma 3 27B and GPT-6 Sol share?
27 benchmarks have published results for both models. Gemma 3 27B has 43 scored results on Noometry and GPT-6 Sol has 45.