Model comparison
Gemma 7B vs Step 3
Step 3 is the stronger model overall, scoring 40.5 to 30.0 on the Noometry Index.
Last verified . 12 shared benchmarks.
Google
StepFunSummary
- They share 12 benchmarks with published results for both. Gemma 7B scores higher in 0 categories and Step 3 in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Step 3 leads 54.3 to 27.1.
Side by side
| Gemma 7B | Step 3 | |
|---|---|---|
| Provider | StepFun | |
| Noometry Index | 30.0 | 40.5 |
| Released | 2024-02-21 | — |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 27 | 17 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Step 3 leads
Gemma 7B: 30.5 (#294), Step 3: 40.1 (#147)
| Benchmark | Gemma 7B | Step 3 |
|---|---|---|
| LMArena Coding | 1048 | 1367 |
| HumanEval+ | 28.7% | — |
| MBPP+ | 43.4% | — |
Reasoning Step 3 leads
Gemma 7B: 19.9 (#249), Step 3: 28.4 (#105)
| Benchmark | Gemma 7B | Step 3 |
|---|---|---|
| LMArena Hard Prompts | 1042 | 1355 |
| Kagi LLM Benchmark | — | 62.3% |
| Adversarial NLI | 48.7% | — |
| BIG-Bench Hard | 55.1% | — |
| Epoch Capabilities Index | 111.99 | — |
| HellaSwag | 82.2% | — |
| PIQA | 81.2% | — |
| WinoGrande | 79% | — |
Math Step 3 leads
Gemma 7B: 31.2 (#228), Step 3: 37.6 (#148)
| Benchmark | Gemma 7B | Step 3 |
|---|---|---|
| LMArena Math | 1066 | 1366 |
| GSM8K | 46.4% | — |
Knowledge Step 3 leads
Gemma 7B: 27.3 (#252), Step 3: 36.8 (#164)
| Benchmark | Gemma 7B | Step 3 |
|---|---|---|
| LMArena Expert | 1001 | 1333 |
| ARC (AI2) Challenge | 78.3% | — |
| BoolQ | 83.2% | — |
| MMLU | 66.1% | — |
| OpenBookQA | 78.6% | — |
| TriviaQA | 72.3% | — |
Multimodal Not comparable
Gemma 7B: —, Step 3: 35.5 (#86)
| Benchmark | Gemma 7B | Step 3 |
|---|---|---|
| LMArena Vision | — | 1177 |
Multilingual Step 3 leads
Gemma 7B: 25.1 (#287), Step 3: 46.3 (#159)
| Benchmark | Gemma 7B | Step 3 |
|---|---|---|
| LMArena Non-English | 999 | 1327 |
| LMArena Chinese | 1035 | 1397 |
| LMArena Russian | 993 | 1331 |
| LMArena French | 1025 | — |
| LMArena German | — | 1371 |
| LMArena Korean | — | 1269 |
| LMArena Spanish | — | 1371 |
Instruction Following Step 3 leads
Gemma 7B: 51.5 (#295), Step 3: 70.4 (#164)
| Benchmark | Gemma 7B | Step 3 |
|---|---|---|
| LMArena Instruction Following | 1017 | 1332 |
Long Context Step 3 leads
Gemma 7B: 31.1 (#282), Step 3: 40.3 (#157)
| Benchmark | Gemma 7B | Step 3 |
|---|---|---|
| LMArena Longer Query | 1022 | 1326 |
Writing & Preference Step 3 leads
Gemma 7B: 27.1 (#302), Step 3: 54.3 (#151)
| Benchmark | Gemma 7B | Step 3 |
|---|---|---|
| LMArena Text | 1056 | 1350 |
| LMArena Creative Writing | 1024 | 1321 |
| LMArena Multi-Turn | 963 | 1341 |
Frequently asked questions
Is Gemma 7B better than Step 3?
Step 3 is the stronger model overall, scoring 40.5 to 30.0 on the Noometry Index.
Is Gemma 7B or Step 3 better for coding?
Step 3 scores higher on coding benchmarks: 40.1 versus 30.5 in the Noometry coding category.
How many benchmarks do Gemma 7B and Step 3 share?
12 benchmarks have published results for both models. Gemma 7B has 27 scored results on Noometry and Step 3 has 17.