Model comparison
Gemma 3 27B vs Llama 2-70B
Gemma 3 27B is the stronger model overall, scoring 30.5 to 24.1 on the Noometry Index.
Last verified . 22 shared benchmarks.
Google
MetaSummary
- They share 22 benchmarks with published results for both. Gemma 3 27B scores higher in 6 categories and Llama 2-70B in 2 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Gemma 3 27B leads 52.7 to 32.5.
- The biggest single-benchmark swing is MATH Level 5: 74% for Gemma 3 27B and 3.3% for Llama 2-70B.
Side by side
| Gemma 3 27B | Llama 2-70B | |
|---|---|---|
| Provider | Meta | |
| Noometry Index | 30.5 | 24.1 |
| Released | 2025-03-11 | 2023-07-18 |
| Weights | Open | Open |
| Context window | 131K | — |
| Max output | 8K | — |
| Input $ / M tokens | $0.08 | — |
| Output $ / M tokens | $0.16 | — |
| Results tracked | 43 | 35 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Llama 2-70B leads
Gemma 3 27B: 22.1 (#329), Llama 2-70B: 30.8 (#281)
| Benchmark | Gemma 3 27B | Llama 2-70B |
|---|---|---|
| LMArena Coding | 1322 | 1079 |
| Aider Polyglot | 4.9% | — |
| SciCode | 21.2% | — |
| LiveBench Coding | 39.9% | — |
Agentic & Tool Use Not comparable
Gemma 3 27B: 24.9 (#105), Llama 2-70B: —
| Benchmark | Gemma 3 27B | Llama 2-70B |
|---|---|---|
| Berkeley Function Calling Leaderboard | 29.5% | — |
Reasoning Gemma 3 27B leads
Gemma 3 27B: 16.3 (#297), Llama 2-70B: 13.7 (#324)
| Benchmark | Gemma 3 27B | Llama 2-70B |
|---|---|---|
| LMArena Hard Prompts | 1340 | 1073 |
| DTBench | 52.5% | 41.6% |
| Epoch Capabilities Index | 130.04 | 113.59 |
| Kagi LLM Benchmark | 40.4% | — |
| CritPt | 0% | — |
| Chess Puzzles | 0% | — |
| LiveBench Reasoning | 43.8% | — |
| LiveBench Data Analysis | 51.5% | — |
| LMCA | 12.3% | — |
| BIG-Bench Hard | — | 64.9% |
| CommonsenseQA 2.0 | — | 50% |
| ForecastBench | — | 51.4 |
| HellaSwag | — | 85.3% |
| LAMBADA | — | 78.9% |
| LiveBench | 50% | — |
| PIQA | — | 82.8% |
| WinoGrande | — | 80.2% |
Math Gemma 3 27B leads
Gemma 3 27B: 24.4 (#261), Llama 2-70B: 7.5 (#320)
| Benchmark | Gemma 3 27B | Llama 2-70B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 22.5% | 0% |
| LMArena Math | 1312 | 1091 |
| MATH Level 5 | 74% | 3.3% |
| LiveBench Math | 55.4% | — |
| GSM8K | — | 69.6% |
Knowledge Gemma 3 27B leads
Gemma 3 27B: 25.4 (#256), Llama 2-70B: 7.4 (#305)
| Benchmark | Gemma 3 27B | Llama 2-70B |
|---|---|---|
| GPQA Diamond | 47.7% | 26.3% |
| LMArena Expert | 1303 | 1039 |
| Confabulations | 40.3% | — |
| Vectara Hallucination Rate | 7.4% | — |
| ARC (AI2) Challenge | — | 78.3% |
| BoolQ | — | 88.6% |
| MMLU | — | 69.9% |
| OpenBookQA | — | 60.2% |
| TriviaQA | — | 87.6% |
Multimodal Not comparable
Gemma 3 27B: 32.0 (#94), Llama 2-70B: —
| Benchmark | Gemma 3 27B | Llama 2-70B |
|---|---|---|
| LMArena Vision | 1165 | — |
| GeoBench | 52% | — |
Multilingual Gemma 3 27B leads
Gemma 3 27B: 47.0 (#150), Llama 2-70B: 27.9 (#269)
| Benchmark | Gemma 3 27B | Llama 2-70B |
|---|---|---|
| LMArena Non-English | 1334 | 1045 |
| LMArena Chinese | 1347 | 994 |
| LMArena French | 1368 | 1089 |
| LMArena German | 1362 | 1041 |
| LMArena Japanese | 1287 | 927 |
| LMArena Korean | 1308 | 964 |
| LMArena Russian | 1349 | 1083 |
| LMArena Spanish | 1348 | 1142 |
Instruction Following Gemma 3 27B leads
Gemma 3 27B: 70.8 (#155), Llama 2-70B: 55.1 (#273)
| Benchmark | Gemma 3 27B | Llama 2-70B |
|---|---|---|
| LMArena Instruction Following | 1321 | 1071 |
| LiveBench Instruction Following | 74.9% | — |
Long Context Llama 2-70B leads
Gemma 3 27B: 27.6 (#288), Llama 2-70B: 32.2 (#265)
| Benchmark | Gemma 3 27B | Llama 2-70B |
|---|---|---|
| LMArena Longer Query | 1333 | 1062 |
| Fiction.LiveBench | 33.3% | — |
Writing & Preference Gemma 3 27B leads
Gemma 3 27B: 52.7 (#161), Llama 2-70B: 32.5 (#274)
| Benchmark | Gemma 3 27B | Llama 2-70B |
|---|---|---|
| LMArena Text | 1358 | 1115 |
| LMArena Creative Writing | 1346 | 1075 |
| LMArena Multi-Turn | 1344 | 1088 |
| Short-Story Creative Writing | 79.9% | — |
| EQ-Bench Creative Writing | 1266 | — |
| LiveBench Language | 34.6% | — |
Frequently asked questions
Is Gemma 3 27B better than Llama 2-70B?
Gemma 3 27B is the stronger model overall, scoring 30.5 to 24.1 on the Noometry Index.
Is Gemma 3 27B or Llama 2-70B better for coding?
Llama 2-70B scores higher on coding benchmarks: 30.8 versus 22.1 in the Noometry coding category.
How many benchmarks do Gemma 3 27B and Llama 2-70B share?
22 benchmarks have published results for both models. Gemma 3 27B has 43 scored results on Noometry and Llama 2-70B has 35.