Model comparison
Deepseek Coder v2 vs Llama 3.1 Nemotron 51b Instruct
Deepseek Coder v2 and Llama 3.1 Nemotron 51b Instruct score almost the same on the Noometry Index (35.9 vs 35.9), so choose on price, context window or the category you care about most.
Last verified . 12 shared benchmarks.
Summary
- They share 12 benchmarks with published results for both. Deepseek Coder v2 scores higher in 6 categories and Llama 3.1 Nemotron 51b Instruct in 2 categories; 3 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Llama 3.1 Nemotron 51b Instruct leads 43.4 to 38.2.
Side by side
| Deepseek Coder v2 | Llama 3.1 Nemotron 51b Instruct | |
|---|---|---|
| Provider | DeepSeek | NVIDIA |
| Noometry Index | 35.9 | 35.9 |
| Released | 2024-06-17 | — |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 24 | 12 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Deepseek Coder v2 leads
Deepseek Coder v2: 38.1 (#183), Llama 3.1 Nemotron 51b Instruct: 35.6 (#222)
| Benchmark | Deepseek Coder v2 | Llama 3.1 Nemotron 51b Instruct |
|---|---|---|
| LMArena Coding | 1251 | 1223 |
| BigCodeBench Instruct | 48.2% | — |
| BigCodeBench Complete | 59.7% | — |
| HumanEval+ | 82.3% | — |
| MBPP+ | 75.1% | — |
Reasoning Too close to call
Deepseek Coder v2: 23.6 (#176), Llama 3.1 Nemotron 51b Instruct: 23.5 (#177)
| Benchmark | Deepseek Coder v2 | Llama 3.1 Nemotron 51b Instruct |
|---|---|---|
| LMArena Hard Prompts | 1207 | 1203 |
| WinoGrande | 83.7% | — |
Math Too close to call
Deepseek Coder v2: 34.9 (#190), Llama 3.1 Nemotron 51b Instruct: 34.6 (#193)
| Benchmark | Deepseek Coder v2 | Llama 3.1 Nemotron 51b Instruct |
|---|---|---|
| LMArena Math | 1241 | 1230 |
| GSM8K | 94.5% | — |
Knowledge Too close to call
Deepseek Coder v2: 32.3 (#212), Llama 3.1 Nemotron 51b Instruct: 31.9 (#218)
| Benchmark | Deepseek Coder v2 | Llama 3.1 Nemotron 51b Instruct |
|---|---|---|
| LMArena Expert | 1181 | 1167 |
| ARC (AI2) Challenge | 64.3% | — |
Multilingual Too close to call
Deepseek Coder v2: 36.3 (#240), Llama 3.1 Nemotron 51b Instruct: 36.1 (#241)
| Benchmark | Deepseek Coder v2 | Llama 3.1 Nemotron 51b Instruct |
|---|---|---|
| LMArena Non-English | 1182 | 1181 |
| LMArena Chinese | 1201 | 1180 |
| LMArena Russian | 1188 | 1187 |
| LMArena French | 1185 | — |
| LMArena German | 1164 | — |
| LMArena Japanese | 1126 | — |
| LMArena Korean | 1104 | — |
| LMArena Spanish | 1153 | — |
Instruction Following Llama 3.1 Nemotron 51b Instruct leads
Deepseek Coder v2: 61.7 (#242), Llama 3.1 Nemotron 51b Instruct: 62.9 (#233)
| Benchmark | Deepseek Coder v2 | Llama 3.1 Nemotron 51b Instruct |
|---|---|---|
| LMArena Instruction Following | 1180 | 1201 |
Long Context Too close to call
Deepseek Coder v2: 37.0 (#224), Llama 3.1 Nemotron 51b Instruct: 36.5 (#230)
| Benchmark | Deepseek Coder v2 | Llama 3.1 Nemotron 51b Instruct |
|---|---|---|
| LMArena Longer Query | 1219 | 1205 |
Writing & Preference Llama 3.1 Nemotron 51b Instruct leads
Deepseek Coder v2: 38.2 (#253), Llama 3.1 Nemotron 51b Instruct: 43.4 (#229)
| Benchmark | Deepseek Coder v2 | Llama 3.1 Nemotron 51b Instruct |
|---|---|---|
| LMArena Text | 1191 | 1228 |
| LMArena Creative Writing | 1120 | 1213 |
| LMArena Multi-Turn | 1177 | 1227 |
Frequently asked questions
Is Deepseek Coder v2 better than Llama 3.1 Nemotron 51b Instruct?
Deepseek Coder v2 and Llama 3.1 Nemotron 51b Instruct score almost the same on the Noometry Index (35.9 vs 35.9), so choose on price, context window or the category you care about most.
Is Deepseek Coder v2 or Llama 3.1 Nemotron 51b Instruct better for coding?
Deepseek Coder v2 scores higher on coding benchmarks: 38.1 versus 35.6 in the Noometry coding category.
How many benchmarks do Deepseek Coder v2 and Llama 3.1 Nemotron 51b Instruct share?
12 benchmarks have published results for both models. Deepseek Coder v2 has 24 scored results on Noometry and Llama 3.1 Nemotron 51b Instruct has 12.