Model comparison
Gemma 2 27B vs o3-pro
o3-pro is the stronger model overall, scoring 42.9 to 29.4 on the Noometry Index. Gemma 2 27B costs 54× less per token, which makes it the better buy when o3-pro's lead doesn't matter for your workload.
Last verified . 4 shared benchmarks.
GoogleSummary
- They share 4 benchmarks with published results for both. Gemma 2 27B scores higher in 0 categories and o3-pro in 5 categories; 5 gaps are clear of the uncertainty.
- The widest gap is in long context, where o3-pro leads 72.2 to 37.3.
- The biggest single-benchmark swing is DTBench: 48% for Gemma 2 27B and 86.9% for o3-pro.
- Gemma 2 27B is cheaper at $0.65 / $0.65 per million input/output tokens, against $20 / $80 for o3-pro.
- o3-pro accepts more context: 200K tokens versus 8K.
- Gemma 2 27B has downloadable open weights; the other is API-only.
Side by side
| Gemma 2 27B | o3-pro | |
|---|---|---|
| Provider | OpenAI | |
| Noometry Index | 29.4 | 42.9 |
| Released | 2024-06-24 | 2025-06-10 |
| Weights | Open | Proprietary |
| Context window | 8K | 200K |
| Max output | 2K | 100K |
| Input $ / M tokens | $0.65 | $20 |
| Output $ / M tokens | $0.65 | $80 |
| Results tracked | 34 | 12 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding o3-pro leads
Gemma 2 27B: 34.1 (#246), o3-pro: 55.5 (#24)
| Benchmark | Gemma 2 27B | o3-pro |
|---|---|---|
| Aider Polyglot | — | 84.9% |
| WeirdML | — | 58.2% |
| BigCodeBench Instruct | 42.8% | — |
| LiveBench Coding | 36% | — |
| LMArena Coding | 1211 | — |
| BigCodeBench Complete | 52.5% | — |
Reasoning o3-pro leads
Gemma 2 27B: 15.3 (#315), o3-pro: 23.8 (#171)
| Benchmark | Gemma 2 27B | o3-pro |
|---|---|---|
| DTBench | 48% | 86.9% |
| LMCA | 7.1% | 38.5% |
| Epoch Capabilities Index | 122.08 | 147.42 |
| ARC-AGI-2 | — | 4.9% |
| Kagi LLM Benchmark | — | 72.1% |
| ARC-AGI-1 | — | 59.3% |
| LiveBench Reasoning | 28.1% | — |
| LMArena Hard Prompts | 1198 | — |
| LiveBench Data Analysis | 47.9% | — |
| LiveBench | 38.2% | — |
Math Not comparable
Gemma 2 27B: 10.7 (#311), o3-pro: —
| Benchmark | Gemma 2 27B | o3-pro |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 1.4% | — |
| LiveBench Math | 26.5% | — |
| LMArena Math | 1212 | — |
| MATH Level 5 | 27.9% | — |
Knowledge o3-pro leads
Gemma 2 27B: 19.0 (#280), o3-pro: 29.5 (#238)
| Benchmark | Gemma 2 27B | o3-pro |
|---|---|---|
| Confabulations | 27.1% | 14.2% |
| GPQA Diamond | 36.5% | — |
| Vectara Hallucination Rate | — | 23.3% |
| LMArena Expert | 1172 | — |
| MMLU | 75.7% | — |
Multilingual Not comparable
Gemma 2 27B: 38.6 (#226), o3-pro: —
| Benchmark | Gemma 2 27B | o3-pro |
|---|---|---|
| LMArena Non-English | 1217 | — |
| LMArena Chinese | 1221 | — |
| LMArena French | 1247 | — |
| LMArena German | 1209 | — |
| LMArena Japanese | 1175 | — |
| LMArena Korean | 1174 | — |
| LMArena Russian | 1234 | — |
| LMArena Spanish | 1228 | — |
Instruction Following Not comparable
Gemma 2 27B: 60.5 (#249), o3-pro: —
| Benchmark | Gemma 2 27B | o3-pro |
|---|---|---|
| LiveBench Instruction Following | 58.1% | — |
| LMArena Instruction Following | 1206 | — |
Long Context o3-pro leads
Gemma 2 27B: 37.3 (#218), o3-pro: 72.2 (#1)
| Benchmark | Gemma 2 27B | o3-pro |
|---|---|---|
| Fiction.LiveBench | — | 97.2% |
| LMArena Longer Query | 1231 | — |
Writing & Preference o3-pro leads
Gemma 2 27B: 44.2 (#225), o3-pro: 57.1 (#133)
| Benchmark | Gemma 2 27B | o3-pro |
|---|---|---|
| LMArena Text | 1231 | — |
| LMArena Creative Writing | 1241 | — |
| Short-Story Creative Writing | — | 84.4% |
| LMArena Multi-Turn | 1224 | — |
| LiveBench Language | 32.6% | — |
Frequently asked questions
Is Gemma 2 27B better than o3-pro?
o3-pro is the stronger model overall, scoring 42.9 to 29.4 on the Noometry Index. Gemma 2 27B costs 54× less per token, which makes it the better buy when o3-pro's lead doesn't matter for your workload.
Which is cheaper, Gemma 2 27B or o3-pro?
Gemma 2 27B is cheaper. It lists at $0.65 per million input tokens and $0.65 per million output tokens; o3-pro lists at $20 and $80.
Is Gemma 2 27B or o3-pro better for coding?
o3-pro scores higher on coding benchmarks: 55.5 versus 34.1 in the Noometry coding category.
Which has the bigger context window?
o3-pro does, with 200K tokens against 8K.
How many benchmarks do Gemma 2 27B and o3-pro share?
4 benchmarks have published results for both models. Gemma 2 27B has 34 scored results on Noometry and o3-pro has 12.