Model comparison
Llama 2-70B vs Step 3
Step 3 is the stronger model overall, scoring 40.5 to 24.4 on the Noometry Index.
Last verified . 15 shared benchmarks.
Meta
StepFunSummary
- They share 15 benchmarks with published results for both. Llama 2-70B 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 8.1.
Side by side
| Llama 2-70B | Step 3 | |
|---|---|---|
| Provider | Meta | StepFun |
| Noometry Index | 24.4 | 40.5 |
| Released | 2023-07-18 | — |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 35 | 17 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Step 3 leads
Llama 2-70B: 31.4 (#286), Step 3: 40.1 (#147)
| Benchmark | Llama 2-70B | Step 3 |
|---|---|---|
| LMArena Coding | 1079 | 1367 |
Reasoning Step 3 leads
Llama 2-70B: 14.4 (#325), Step 3: 28.4 (#105)
| Benchmark | Llama 2-70B | Step 3 |
|---|---|---|
| LMArena Hard Prompts | 1073 | 1355 |
| Kagi LLM Benchmark | — | 62.3% |
| DTBench | 41.6% | — |
| BIG-Bench Hard | 64.9% | — |
| CommonsenseQA 2.0 | 50% | — |
| Epoch Capabilities Index | 113.79 | — |
| ForecastBench | 51.4 | — |
| HellaSwag | 85.3% | — |
| LAMBADA | 78.9% | — |
| PIQA | 82.8% | — |
| WinoGrande | 80.2% | — |
Math Step 3 leads
Llama 2-70B: 8.1 (#326), Step 3: 37.6 (#148)
| Benchmark | Llama 2-70B | Step 3 |
|---|---|---|
| LMArena Math | 1091 | 1366 |
| OTIS Mock AIME 2024-2025 | 0% | — |
| MATH Level 5 | 3.3% | — |
| GSM8K | 69.6% | — |
Knowledge Step 3 leads
Llama 2-70B: 7.4 (#310), Step 3: 36.8 (#164)
| Benchmark | Llama 2-70B | Step 3 |
|---|---|---|
| LMArena Expert | 1039 | 1333 |
| GPQA Diamond | 26.3% | — |
| ARC (AI2) Challenge | 78.3% | — |
| BoolQ | 88.6% | — |
| MMLU | 69.9% | — |
| OpenBookQA | 60.2% | — |
| TriviaQA | 87.6% | — |
Multimodal Not comparable
Llama 2-70B: —, Step 3: 35.5 (#86)
| Benchmark | Llama 2-70B | Step 3 |
|---|---|---|
| LMArena Vision | — | 1177 |
Multilingual Step 3 leads
Llama 2-70B: 27.7 (#274), Step 3: 46.3 (#159)
| Benchmark | Llama 2-70B | Step 3 |
|---|---|---|
| LMArena Non-English | 1045 | 1327 |
| LMArena Chinese | 995 | 1397 |
| LMArena German | 1041 | 1371 |
| LMArena Korean | 964 | 1269 |
| LMArena Russian | 1083 | 1331 |
| LMArena Spanish | 1143 | 1371 |
| LMArena French | 1090 | — |
| LMArena Japanese | 927 | — |
Instruction Following Step 3 leads
Llama 2-70B: 54.9 (#278), Step 3: 70.4 (#164)
| Benchmark | Llama 2-70B | Step 3 |
|---|---|---|
| LMArena Instruction Following | 1071 | 1332 |
Long Context Step 3 leads
Llama 2-70B: 32.3 (#270), Step 3: 40.3 (#157)
| Benchmark | Llama 2-70B | Step 3 |
|---|---|---|
| LMArena Longer Query | 1062 | 1326 |
Writing & Preference Step 3 leads
Llama 2-70B: 32.3 (#279), Step 3: 54.3 (#151)
| Benchmark | Llama 2-70B | Step 3 |
|---|---|---|
| LMArena Text | 1115 | 1350 |
| LMArena Creative Writing | 1075 | 1321 |
| LMArena Multi-Turn | 1088 | 1341 |
Frequently asked questions
Is Llama 2-70B better than Step 3?
Step 3 is the stronger model overall, scoring 40.5 to 24.4 on the Noometry Index.
Is Llama 2-70B or Step 3 better for coding?
Step 3 scores higher on coding benchmarks: 40.1 versus 31.4 in the Noometry coding category.
How many benchmarks do Llama 2-70B and Step 3 share?
15 benchmarks have published results for both models. Llama 2-70B has 35 scored results on Noometry and Step 3 has 17.