Model comparison
Llama 3-70B vs StarCoder 2 15B
Llama 3-70B has enough public results to be ranked (#323); StarCoder 2 15B does not yet, so treat this comparison as directional.
Last verified . 5 shared benchmarks.
Meta
NVIDIASummary
- They share 5 benchmarks with published results for both. Llama 3-70B scores higher in 0 categories and StarCoder 2 15B in 1 category, but none of those gaps is larger than the uncertainty.
- The biggest single-benchmark swing is BigCodeBench Complete: 54.5% for Llama 3-70B and 38.4% for StarCoder 2 15B.
Side by side
| Llama 3-70B | StarCoder 2 15B | |
|---|---|---|
| Provider | Meta | NVIDIA |
| Noometry Index | 28.8 | 38.1 |
| Released | 2024-04-18 | 2024-02-20 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 31 | 7 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
Llama 3-70B: 35.8 (#218), StarCoder 2 15B: 36.6
| Benchmark | Llama 3-70B | StarCoder 2 15B |
|---|---|---|
| BigCodeBench Complete | 54.5% | 38.4% |
| HumanEval+ | 72% | 37.8% |
| BigCodeBench Instruct | 43.6% | — |
| LMArena Coding | 1206 | — |
| MBPP+ | 69% | — |
Agentic & Tool Use Not comparable
Llama 3-70B: 21.1 (#139), StarCoder 2 15B: —
| Benchmark | Llama 3-70B | StarCoder 2 15B |
|---|---|---|
| Cybench | 5% | — |
Reasoning Not comparable
Llama 3-70B: 18.0 (#288), StarCoder 2 15B: —
| Benchmark | Llama 3-70B | StarCoder 2 15B |
|---|---|---|
| Epoch Capabilities Index | 122.93 | 105 |
| WinoGrande | 83.5% | 64.3% |
| Kagi LLM Benchmark | 35.1% | — |
| LMArena Hard Prompts | 1195 | — |
| DTBench | 54.2% | — |
| ForecastBench | 57.1 | — |
Math Not comparable
Llama 3-70B: 12.8 (#305), StarCoder 2 15B: —
| Benchmark | Llama 3-70B | StarCoder 2 15B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 4.3% | — |
| LMArena Math | 1218 | — |
| MATH Level 5 | 22.6% | — |
| GSM8K | — | 57.7% |
Knowledge Not comparable
Llama 3-70B: 20.8 (#277), StarCoder 2 15B: —
| Benchmark | Llama 3-70B | StarCoder 2 15B |
|---|---|---|
| MMLU | 79.3% | 64.1% |
| GPQA Diamond | 40.6% | — |
| LMArena Expert | 1149 | — |
| ARC (AI2) Challenge | — | 47.2% |
Multilingual Not comparable
Llama 3-70B: 33.6 (#251), StarCoder 2 15B: —
| Benchmark | Llama 3-70B | StarCoder 2 15B |
|---|---|---|
| LMArena Non-English | 1142 | — |
| LMArena Chinese | 1114 | — |
| LMArena French | 1232 | — |
| LMArena German | 1169 | — |
| LMArena Japanese | 1017 | — |
| LMArena Korean | 1017 | — |
| LMArena Russian | 1159 | — |
| LMArena Spanish | 1241 | — |
Instruction Following Not comparable
Llama 3-70B: 62.5 (#238), StarCoder 2 15B: —
| Benchmark | Llama 3-70B | StarCoder 2 15B |
|---|---|---|
| LMArena Instruction Following | 1194 | — |
Long Context Not comparable
Llama 3-70B: 35.6 (#240), StarCoder 2 15B: —
| Benchmark | Llama 3-70B | StarCoder 2 15B |
|---|---|---|
| LMArena Longer Query | 1174 | — |
Writing & Preference Not comparable
Llama 3-70B: 42.8 (#231), StarCoder 2 15B: —
| Benchmark | Llama 3-70B | StarCoder 2 15B |
|---|---|---|
| LMArena Text | 1221 | — |
| LMArena Creative Writing | 1210 | — |
| LMArena Multi-Turn | 1223 | — |
Frequently asked questions
Is Llama 3-70B better than StarCoder 2 15B?
Llama 3-70B has enough public results to be ranked (#323); StarCoder 2 15B does not yet, so treat this comparison as directional.
Is Llama 3-70B or StarCoder 2 15B better for coding?
They score almost the same on coding (35.8 vs 36.6); test both on your own repository before choosing.
How many benchmarks do Llama 3-70B and StarCoder 2 15B share?
5 benchmarks have published results for both models. Llama 3-70B has 31 scored results on Noometry and StarCoder 2 15B has 7.