Model comparison
Llama 3-70B vs Step 3
Step 3 is the stronger model overall, scoring 40.5 to 28.8 on the Noometry Index.
Last verified . 16 shared benchmarks.
Meta
StepFunSummary
- They share 16 benchmarks with published results for both. Llama 3-70B 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 12.8.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 35.1% for Llama 3-70B and 62.3% for Step 3.
Side by side
| Llama 3-70B | Step 3 | |
|---|---|---|
| Provider | Meta | StepFun |
| Noometry Index | 28.8 | 40.5 |
| Released | 2024-04-18 | — |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 31 | 17 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Step 3 leads
Llama 3-70B: 35.8 (#218), Step 3: 40.1 (#147)
| Benchmark | Llama 3-70B | Step 3 |
|---|---|---|
| LMArena Coding | 1206 | 1367 |
| BigCodeBench Instruct | 43.6% | — |
| BigCodeBench Complete | 54.5% | — |
| HumanEval+ | 72% | — |
| MBPP+ | 69% | — |
Agentic & Tool Use Not comparable
Llama 3-70B: 21.1 (#139), Step 3: —
| Benchmark | Llama 3-70B | Step 3 |
|---|---|---|
| Cybench | 5% | — |
Reasoning Step 3 leads
Llama 3-70B: 18.0 (#288), Step 3: 28.4 (#105)
| Benchmark | Llama 3-70B | Step 3 |
|---|---|---|
| Kagi LLM Benchmark | 35.1% | 62.3% |
| LMArena Hard Prompts | 1195 | 1355 |
| DTBench | 54.2% | — |
| Epoch Capabilities Index | 122.93 | — |
| ForecastBench | 57.1 | — |
| WinoGrande | 83.5% | — |
Math Step 3 leads
Llama 3-70B: 12.8 (#305), Step 3: 37.6 (#148)
| Benchmark | Llama 3-70B | Step 3 |
|---|---|---|
| LMArena Math | 1218 | 1366 |
| OTIS Mock AIME 2024-2025 | 4.3% | — |
| MATH Level 5 | 22.6% | — |
Knowledge Step 3 leads
Llama 3-70B: 20.8 (#277), Step 3: 36.8 (#164)
| Benchmark | Llama 3-70B | Step 3 |
|---|---|---|
| LMArena Expert | 1149 | 1333 |
| GPQA Diamond | 40.6% | — |
| MMLU | 79.3% | — |
Multimodal Not comparable
Llama 3-70B: —, Step 3: 35.5 (#86)
| Benchmark | Llama 3-70B | Step 3 |
|---|---|---|
| LMArena Vision | — | 1177 |
Multilingual Step 3 leads
Llama 3-70B: 33.6 (#251), Step 3: 46.3 (#159)
| Benchmark | Llama 3-70B | Step 3 |
|---|---|---|
| LMArena Non-English | 1142 | 1327 |
| LMArena Chinese | 1114 | 1397 |
| LMArena German | 1169 | 1371 |
| LMArena Korean | 1017 | 1269 |
| LMArena Russian | 1159 | 1331 |
| LMArena Spanish | 1241 | 1371 |
| LMArena French | 1232 | — |
| LMArena Japanese | 1017 | — |
Instruction Following Step 3 leads
Llama 3-70B: 62.5 (#238), Step 3: 70.4 (#164)
| Benchmark | Llama 3-70B | Step 3 |
|---|---|---|
| LMArena Instruction Following | 1194 | 1332 |
Long Context Step 3 leads
Llama 3-70B: 35.6 (#240), Step 3: 40.3 (#157)
| Benchmark | Llama 3-70B | Step 3 |
|---|---|---|
| LMArena Longer Query | 1174 | 1326 |
Writing & Preference Step 3 leads
Llama 3-70B: 42.8 (#231), Step 3: 54.3 (#151)
| Benchmark | Llama 3-70B | Step 3 |
|---|---|---|
| LMArena Text | 1221 | 1350 |
| LMArena Creative Writing | 1210 | 1321 |
| LMArena Multi-Turn | 1223 | 1341 |
Frequently asked questions
Is Llama 3-70B better than Step 3?
Step 3 is the stronger model overall, scoring 40.5 to 28.8 on the Noometry Index.
Is Llama 3-70B or Step 3 better for coding?
Step 3 scores higher on coding benchmarks: 40.1 versus 35.8 in the Noometry coding category.
How many benchmarks do Llama 3-70B and Step 3 share?
16 benchmarks have published results for both models. Llama 3-70B has 31 scored results on Noometry and Step 3 has 17.