Model comparison
Gemma 2 27B vs Llama 4 Maverick
Llama 4 Maverick is the stronger model overall, scoring 30.9 to 29.4 on the Noometry Index.
Last verified . 26 shared benchmarks.
Google
MetaSummary
- They share 26 benchmarks with published results for both. Gemma 2 27B scores higher in 4 categories and Llama 4 Maverick in 4 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Llama 4 Maverick leads 26.0 to 10.7.
- The biggest single-benchmark swing is MATH Level 5: 27.9% for Gemma 2 27B and 73% for Llama 4 Maverick.
- Llama 4 Maverick is cheaper at $0.19 / $0.65 per million input/output tokens, against $0.65 / $0.65 for Gemma 2 27B.
- Llama 4 Maverick accepts more context: 128K tokens versus 8K.
Side by side
| Gemma 2 27B | Llama 4 Maverick | |
|---|---|---|
| Provider | Meta | |
| Noometry Index | 29.4 | 30.9 |
| Released | 2024-06-24 | 2025-04-05 |
| Weights | Open | Open |
| Context window | 8K | 128K |
| Max output | 2K | 4K |
| Input $ / M tokens | $0.65 | $0.19 |
| Output $ / M tokens | $0.65 | $0.65 |
| Results tracked | 34 | 54 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Gemma 2 27B leads
Gemma 2 27B: 34.1 (#246), Llama 4 Maverick: 26.6 (#324)
| Benchmark | Gemma 2 27B | Llama 4 Maverick |
|---|---|---|
| BigCodeBench Instruct | 42.8% | 49.7% |
| LMArena Coding | 1211 | 1302 |
| BigCodeBench Complete | 52.5% | 61.4% |
| SWE-bench Verified (bash only) | — | 21% |
| Aider Polyglot | — | 15.6% |
| SciCode | — | 33.1% |
| WeirdML | — | 24.5% |
| LiveBench Coding | 36% | — |
| ALE-Bench | — | 172.97 |
Agentic & Tool Use Not comparable
Gemma 2 27B: —, Llama 4 Maverick: 28.2 (#91)
| Benchmark | Gemma 2 27B | Llama 4 Maverick |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 37.3% |
Reasoning Gemma 2 27B leads
Gemma 2 27B: 15.3 (#315), Llama 4 Maverick: 10.1 (#342)
| Benchmark | Gemma 2 27B | Llama 4 Maverick |
|---|---|---|
| LMArena Hard Prompts | 1198 | 1281 |
| DTBench | 48% | 61.9% |
| LMCA | 7.1% | 15.9% |
| Epoch Capabilities Index | 122.08 | 132.2 |
| ARC-AGI-2 | — | 0% |
| SimpleBench | — | 27.7% |
| Kagi LLM Benchmark | — | 55.9% |
| NYT Connections (extended) | — | 8% |
| ARC-AGI-1 | — | 4.4% |
| CritPt | — | 0% |
| EnigmaEval | — | 0.6% |
| LiveBench Reasoning | 28.1% | — |
| LiveBench Data Analysis | 47.9% | — |
| ForecastBench | — | 57.5 |
| LiveBench | 38.2% | — |
Math Llama 4 Maverick leads
Gemma 2 27B: 10.7 (#311), Llama 4 Maverick: 26.0 (#262)
| Benchmark | Gemma 2 27B | Llama 4 Maverick |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 1.4% | 20.6% |
| LMArena Math | 1212 | 1299 |
| MATH Level 5 | 27.9% | 73% |
| Omni-MATH | — | 42.2% |
| LiveBench Math | 26.5% | — |
| FrontierMath (Feb 2025 set) | — | 0.7% |
Knowledge Llama 4 Maverick leads
Gemma 2 27B: 19.0 (#280), Llama 4 Maverick: 33.4 (#204)
| Benchmark | Gemma 2 27B | Llama 4 Maverick |
|---|---|---|
| GPQA Diamond | 36.5% | 67% |
| Confabulations | 27.1% | 22.6% |
| LMArena Expert | 1172 | 1259 |
| Humanity's Last Exam | — | 5.7% |
| MMLU-Pro | — | 81% |
| Vectara Hallucination Rate | — | 8.2% |
| GPQA (HELM) | — | 65% |
| MMLU | 75.7% | — |
Multimodal Not comparable
Gemma 2 27B: —, Llama 4 Maverick: 31.6 (#105)
| Benchmark | Gemma 2 27B | Llama 4 Maverick |
|---|---|---|
| LMArena Vision | — | 1142 |
| GeoBench | — | 52% |
| SpatialViz-Bench | — | 31.8% |
Multilingual Llama 4 Maverick leads
Gemma 2 27B: 38.6 (#226), Llama 4 Maverick: 42.2 (#195)
| Benchmark | Gemma 2 27B | Llama 4 Maverick |
|---|---|---|
| LMArena Non-English | 1217 | 1269 |
| LMArena Chinese | 1221 | 1277 |
| LMArena French | 1247 | 1259 |
| LMArena German | 1209 | 1291 |
| LMArena Japanese | 1175 | 1207 |
| LMArena Korean | 1174 | 1203 |
| LMArena Russian | 1234 | 1286 |
| LMArena Spanish | 1228 | 1293 |
Instruction Following Llama 4 Maverick leads
Gemma 2 27B: 60.5 (#249), Llama 4 Maverick: 71.7 (#146)
| Benchmark | Gemma 2 27B | Llama 4 Maverick |
|---|---|---|
| LMArena Instruction Following | 1206 | 1267 |
| LiveBench Instruction Following | 58.1% | — |
| IFEval | — | 90.8% |
Long Context Gemma 2 27B leads
Gemma 2 27B: 37.3 (#218), Llama 4 Maverick: 31.4 (#279)
| Benchmark | Gemma 2 27B | Llama 4 Maverick |
|---|---|---|
| LMArena Longer Query | 1231 | 1280 |
| Fiction.LiveBench | — | 46.2% |
Writing & Preference Gemma 2 27B leads
Gemma 2 27B: 44.2 (#225), Llama 4 Maverick: 38.8 (#252)
| Benchmark | Gemma 2 27B | Llama 4 Maverick |
|---|---|---|
| LMArena Text | 1231 | 1287 |
| LMArena Creative Writing | 1241 | 1267 |
| LMArena Multi-Turn | 1224 | 1289 |
| Short-Story Creative Writing | — | 62% |
| EQ-Bench Creative Writing | — | 860 |
| WildBench | — | 80% |
| LiveBench Language | 32.6% | — |
Frequently asked questions
Is Gemma 2 27B better than Llama 4 Maverick?
Llama 4 Maverick is the stronger model overall, scoring 30.9 to 29.4 on the Noometry Index.
Which is cheaper, Gemma 2 27B or Llama 4 Maverick?
Llama 4 Maverick is cheaper. It lists at $0.19 per million input tokens and $0.65 per million output tokens; Gemma 2 27B lists at $0.65 and $0.65.
Is Gemma 2 27B or Llama 4 Maverick better for coding?
Gemma 2 27B scores higher on coding benchmarks: 34.1 versus 26.6 in the Noometry coding category.
Which has the bigger context window?
Llama 4 Maverick does, with 128K tokens against 8K.
How many benchmarks do Gemma 2 27B and Llama 4 Maverick share?
26 benchmarks have published results for both models. Gemma 2 27B has 34 scored results on Noometry and Llama 4 Maverick has 54.