Model comparison
Gemma 3 27B vs Llama 4 Scout
Gemma 3 27B is the stronger model overall, scoring 30.8 to 27.7 on the Noometry Index.
Last verified . 31 shared benchmarks.
Google
MetaSummary
- They share 31 benchmarks with published results for both. Gemma 3 27B scores higher in 9 categories and Llama 4 Scout in 1 category; 7 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Gemma 3 27B leads 52.5 to 37.0.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 22.5% for Gemma 3 27B and 7.8% for Llama 4 Scout.
- Gemma 3 27B is cheaper at $0.08 / $0.16 per million input/output tokens, against $0.10 / $0.30 for Llama 4 Scout.
- Gemma 3 27B accepts more context: 131K tokens versus 128K.
Side by side
| Gemma 3 27B | Llama 4 Scout | |
|---|---|---|
| Provider | Meta | |
| Noometry Index | 30.8 | 27.7 |
| Released | 2025-03-11 | 2025-04-05 |
| Weights | Open | Open |
| Context window | 131K | 128K |
| Max output | 8K | 4K |
| Input $ / M tokens | $0.08 | $0.10 |
| Output $ / M tokens | $0.16 | $0.30 |
| Results tracked | 43 | 43 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Gemma 3 27B leads
Gemma 3 27B: 22.5 (#334), Llama 4 Scout: 20.2 (#339)
| Benchmark | Gemma 3 27B | Llama 4 Scout |
|---|---|---|
| SciCode | 21.2% | 17% |
| LMArena Coding | 1322 | 1286 |
| SWE-bench Verified (bash only) | — | 9.1% |
| Aider Polyglot | 4.9% | — |
| LiveBench Coding | 39.9% | — |
| BigCodeBench Complete | — | 43.1% |
Agentic & Tool Use Too close to call
Gemma 3 27B: 25.1 (#110), Llama 4 Scout: 24.6 (#119)
| Benchmark | Gemma 3 27B | Llama 4 Scout |
|---|---|---|
| Berkeley Function Calling Leaderboard | 29.5% | 28.1% |
Reasoning Gemma 3 27B leads
Gemma 3 27B: 16.7 (#301), Llama 4 Scout: 9.1 (#345)
| Benchmark | Gemma 3 27B | Llama 4 Scout |
|---|---|---|
| Kagi LLM Benchmark | 40.4% | 36.9% |
| CritPt | 0% | 0% |
| LMArena Hard Prompts | 1340 | 1266 |
| DTBench | 52.5% | 57.9% |
| LMCA | 12.3% | 12% |
| Epoch Capabilities Index | 130.04 | 129.64 |
| ARC-AGI-2 | — | 0% |
| ARC-AGI-1 | — | 0.5% |
| Chess Puzzles | 0% | — |
| LiveBench Reasoning | 43.8% | — |
| LiveBench Data Analysis | 51.5% | — |
| ForecastBench | — | 57.5 |
| LiveBench | 50% | — |
Math Gemma 3 27B leads
Gemma 3 27B: 25.9 (#265), Llama 4 Scout: 19.6 (#286)
| Benchmark | Gemma 3 27B | Llama 4 Scout |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 22.5% | 7.8% |
| LMArena Math | 1312 | 1287 |
| MATH Level 5 | 74% | 62.3% |
| Omni-MATH | — | 37.3% |
| LiveBench Math | 55.4% | — |
| FrontierMath (Feb 2025 set) | — | 0% |
Knowledge Llama 4 Scout leads
Gemma 3 27B: 25.5 (#261), Llama 4 Scout: 31.9 (#217)
| Benchmark | Gemma 3 27B | Llama 4 Scout |
|---|---|---|
| GPQA Diamond | 47.7% | 51.8% |
| Vectara Hallucination Rate | 7.4% | 7.7% |
| LMArena Expert | 1304 | 1235 |
| MMLU-Pro | — | 74.2% |
| Confabulations | 40.3% | — |
| GPQA (HELM) | — | 50.7% |
Multimodal Too close to call
Gemma 3 27B: 32.6 (#100), Llama 4 Scout: 32.2 (#102)
| Benchmark | Gemma 3 27B | Llama 4 Scout |
|---|---|---|
| LMArena Vision | 1164 | 1118 |
| GeoBench | 52% | — |
| SpatialViz-Bench | — | 34.2% |
Multilingual Gemma 3 27B leads
Gemma 3 27B: 46.9 (#155), Llama 4 Scout: 41.0 (#212)
| Benchmark | Gemma 3 27B | Llama 4 Scout |
|---|---|---|
| LMArena Non-English | 1334 | 1252 |
| LMArena Chinese | 1346 | 1255 |
| LMArena French | 1368 | 1282 |
| LMArena German | 1362 | 1272 |
| LMArena Japanese | 1287 | 1206 |
| LMArena Korean | 1308 | 1207 |
| LMArena Russian | 1349 | 1263 |
| LMArena Spanish | 1349 | 1278 |
Instruction Following Gemma 3 27B leads
Gemma 3 27B: 70.6 (#160), Llama 4 Scout: 65.8 (#217)
| Benchmark | Gemma 3 27B | Llama 4 Scout |
|---|---|---|
| LMArena Instruction Following | 1321 | 1248 |
| LiveBench Instruction Following | 74.9% | — |
| IFEval | — | 81.8% |
Long Context Too close to call
Gemma 3 27B: 27.6 (#293), Llama 4 Scout: 27.5 (#294)
| Benchmark | Gemma 3 27B | Llama 4 Scout |
|---|---|---|
| Fiction.LiveBench | 33.3% | 36% |
| LMArena Longer Query | 1333 | 1265 |
Writing & Preference Gemma 3 27B leads
Gemma 3 27B: 52.5 (#168), Llama 4 Scout: 37.0 (#261)
| Benchmark | Gemma 3 27B | Llama 4 Scout |
|---|---|---|
| LMArena Text | 1358 | 1279 |
| LMArena Creative Writing | 1346 | 1249 |
| EQ-Bench Creative Writing | 1266 | 783 |
| LMArena Multi-Turn | 1345 | 1280 |
| Short-Story Creative Writing | 79.9% | — |
| WildBench | — | 78% |
| LiveBench Language | 34.6% | — |
Frequently asked questions
Is Gemma 3 27B better than Llama 4 Scout?
Gemma 3 27B is the stronger model overall, scoring 30.8 to 27.7 on the Noometry Index.
Which is cheaper, Gemma 3 27B or Llama 4 Scout?
Gemma 3 27B is cheaper. It lists at $0.08 per million input tokens and $0.16 per million output tokens; Llama 4 Scout lists at $0.10 and $0.30.
Is Gemma 3 27B or Llama 4 Scout better for coding?
Gemma 3 27B scores higher on coding benchmarks: 22.5 versus 20.2 in the Noometry coding category.
Which has the bigger context window?
Gemma 3 27B does, with 131K tokens against 128K.
How many benchmarks do Gemma 3 27B and Llama 4 Scout share?
31 benchmarks have published results for both models. Gemma 3 27B has 43 scored results on Noometry and Llama 4 Scout has 43.