Model comparison
Gemma 2 27B vs Step 3
Step 3 is the stronger model overall, scoring 40.5 to 29.4 on the Noometry Index.
Last verified . 15 shared benchmarks.
Google
StepFunSummary
- They share 15 benchmarks with published results for both. Gemma 2 27B scores higher in 0 categories and Step 3 in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Step 3 leads 37.6 to 10.7.
Side by side
| Gemma 2 27B | Step 3 | |
|---|---|---|
| Provider | StepFun | |
| Noometry Index | 29.4 | 40.5 |
| Released | 2024-06-24 | — |
| Weights | Open | Open |
| Context window | 8K | — |
| Max output | 2K | — |
| Input $ / M tokens | $0.65 | — |
| Output $ / M tokens | $0.65 | — |
| Results tracked | 34 | 17 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Step 3 leads
Gemma 2 27B: 34.1 (#246), Step 3: 40.1 (#147)
| Benchmark | Gemma 2 27B | Step 3 |
|---|---|---|
| LMArena Coding | 1211 | 1367 |
| BigCodeBench Instruct | 42.8% | — |
| LiveBench Coding | 36% | — |
| BigCodeBench Complete | 52.5% | — |
Reasoning Step 3 leads
Gemma 2 27B: 15.3 (#315), Step 3: 28.4 (#105)
| Benchmark | Gemma 2 27B | Step 3 |
|---|---|---|
| LMArena Hard Prompts | 1198 | 1355 |
| Kagi LLM Benchmark | — | 62.3% |
| LiveBench Reasoning | 28.1% | — |
| DTBench | 48% | — |
| LiveBench Data Analysis | 47.9% | — |
| LMCA | 7.1% | — |
| Epoch Capabilities Index | 122.08 | — |
| LiveBench | 38.2% | — |
Math Step 3 leads
Gemma 2 27B: 10.7 (#311), Step 3: 37.6 (#148)
| Benchmark | Gemma 2 27B | Step 3 |
|---|---|---|
| LMArena Math | 1212 | 1366 |
| OTIS Mock AIME 2024-2025 | 1.4% | — |
| LiveBench Math | 26.5% | — |
| MATH Level 5 | 27.9% | — |
Knowledge Step 3 leads
Gemma 2 27B: 19.0 (#280), Step 3: 36.8 (#164)
| Benchmark | Gemma 2 27B | Step 3 |
|---|---|---|
| LMArena Expert | 1172 | 1333 |
| GPQA Diamond | 36.5% | — |
| Confabulations | 27.1% | — |
| MMLU | 75.7% | — |
Multimodal Not comparable
Gemma 2 27B: —, Step 3: 35.5 (#86)
| Benchmark | Gemma 2 27B | Step 3 |
|---|---|---|
| LMArena Vision | — | 1177 |
Multilingual Step 3 leads
Gemma 2 27B: 38.6 (#226), Step 3: 46.3 (#159)
| Benchmark | Gemma 2 27B | Step 3 |
|---|---|---|
| LMArena Non-English | 1217 | 1327 |
| LMArena Chinese | 1221 | 1397 |
| LMArena German | 1209 | 1371 |
| LMArena Korean | 1174 | 1269 |
| LMArena Russian | 1234 | 1331 |
| LMArena Spanish | 1228 | 1371 |
| LMArena French | 1247 | — |
| LMArena Japanese | 1175 | — |
Instruction Following Step 3 leads
Gemma 2 27B: 60.5 (#249), Step 3: 70.4 (#164)
| Benchmark | Gemma 2 27B | Step 3 |
|---|---|---|
| LMArena Instruction Following | 1206 | 1332 |
| LiveBench Instruction Following | 58.1% | — |
Long Context Step 3 leads
Gemma 2 27B: 37.3 (#218), Step 3: 40.3 (#157)
| Benchmark | Gemma 2 27B | Step 3 |
|---|---|---|
| LMArena Longer Query | 1231 | 1326 |
Writing & Preference Step 3 leads
Gemma 2 27B: 44.2 (#225), Step 3: 54.3 (#151)
| Benchmark | Gemma 2 27B | Step 3 |
|---|---|---|
| LMArena Text | 1231 | 1350 |
| LMArena Creative Writing | 1241 | 1321 |
| LMArena Multi-Turn | 1224 | 1341 |
| LiveBench Language | 32.6% | — |
Frequently asked questions
Is Gemma 2 27B better than Step 3?
Step 3 is the stronger model overall, scoring 40.5 to 29.4 on the Noometry Index.
Is Gemma 2 27B or Step 3 better for coding?
Step 3 scores higher on coding benchmarks: 40.1 versus 34.1 in the Noometry coding category.
How many benchmarks do Gemma 2 27B and Step 3 share?
15 benchmarks have published results for both models. Gemma 2 27B has 34 scored results on Noometry and Step 3 has 17.