Model comparison
DeepSeek-V3.1 vs Llama 2-13B
DeepSeek-V3.1 is the stronger model overall, scoring 42.8 to 29.6 on the Noometry Index.
Last verified . 19 shared benchmarks.
DeepSeek
MetaSummary
- They share 19 benchmarks with published results for both. DeepSeek-V3.1 scores higher in 8 categories and Llama 2-13B in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where DeepSeek-V3.1 leads 60.3 to 29.8.
- The biggest single-benchmark swing is DTBench: 82.7% for DeepSeek-V3.1 and 42.2% for Llama 2-13B.
Side by side
| DeepSeek-V3.1 | Llama 2-13B | |
|---|---|---|
| Provider | DeepSeek | Meta |
| Noometry Index | 42.8 | 29.6 |
| Released | 2025-08-21 | 2023-07-18 |
| Weights | Open | Open |
| Context window | 164K | — |
| Max output | 8K | — |
| Input $ / M tokens | $0.25 | — |
| Output $ / M tokens | $0.95 | — |
| Results tracked | 27 | 32 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek-V3.1 leads
DeepSeek-V3.1: 40.3 (#144), Llama 2-13B: 30.9 (#291)
| Benchmark | DeepSeek-V3.1 | Llama 2-13B |
|---|---|---|
| LMArena Coding | 1417 | 1062 |
| WeirdML | 38.4% | — |
Reasoning DeepSeek-V3.1 leads
DeepSeek-V3.1: 27.9 (#110), Llama 2-13B: 12.8 (#337)
| Benchmark | DeepSeek-V3.1 | Llama 2-13B |
|---|---|---|
| LMArena Hard Prompts | 1417 | 1051 |
| DTBench | 82.7% | 42.2% |
| Epoch Capabilities Index | 139.92 | 106.17 |
| SimpleBench | 40% | — |
| Kagi LLM Benchmark | 53.2% | — |
| Chess Puzzles | — | 0% |
| LMCA | 24.3% | — |
| BIG-Bench Hard | — | 58.2% |
| ForecastBench | 58 | — |
| HellaSwag | — | 80.7% |
| LAMBADA | — | 76.5% |
| PIQA | — | 80.8% |
| WinoGrande | — | 72.8% |
Math DeepSeek-V3.1 leads
DeepSeek-V3.1: 38.9 (#122), Llama 2-13B: 31.1 (#229)
| Benchmark | DeepSeek-V3.1 | Llama 2-13B |
|---|---|---|
| LMArena Math | 1420 | 1065 |
| GSM8K | — | 36.9% |
Knowledge DeepSeek-V3.1 leads
DeepSeek-V3.1: 43.7 (#90), Llama 2-13B: 28.1 (#249)
| Benchmark | DeepSeek-V3.1 | Llama 2-13B |
|---|---|---|
| LMArena Expert | 1405 | 1030 |
| Vectara Hallucination Rate | 5.5% | — |
| ARC (AI2) Challenge | — | 60.3% |
| BoolQ | — | 82.4% |
| MMLU | — | 55.6% |
| OpenBookQA | — | 57% |
| TriviaQA | — | 79.6% |
Multimodal Not comparable
DeepSeek-V3.1: —, Llama 2-13B: —
| Benchmark | DeepSeek-V3.1 | Llama 2-13B |
|---|---|---|
| ScienceQA | — | 55.8% |
Multilingual DeepSeek-V3.1 leads
DeepSeek-V3.1: 51.6 (#106), Llama 2-13B: 26.5 (#279)
| Benchmark | DeepSeek-V3.1 | Llama 2-13B |
|---|---|---|
| LMArena Non-English | 1400 | 1024 |
| LMArena Chinese | 1469 | 1001 |
| LMArena French | 1447 | 1044 |
| LMArena German | 1411 | 1009 |
| LMArena Japanese | 1378 | 894 |
| LMArena Korean | 1337 | 953 |
| LMArena Russian | 1405 | 1055 |
| LMArena Spanish | 1431 | 1087 |
Instruction Following DeepSeek-V3.1 leads
DeepSeek-V3.1: 73.9 (#110), Llama 2-13B: 53.3 (#287)
| Benchmark | DeepSeek-V3.1 | Llama 2-13B |
|---|---|---|
| LMArena Instruction Following | 1400 | 1045 |
Long Context DeepSeek-V3.1 leads
DeepSeek-V3.1: 36.3 (#232), Llama 2-13B: 32.3 (#269)
| Benchmark | DeepSeek-V3.1 | Llama 2-13B |
|---|---|---|
| LMArena Longer Query | 1422 | 1064 |
| Fiction.LiveBench | 52.8% | — |
Writing & Preference DeepSeek-V3.1 leads
DeepSeek-V3.1: 60.3 (#98), Llama 2-13B: 29.8 (#289)
| Benchmark | DeepSeek-V3.1 | Llama 2-13B |
|---|---|---|
| LMArena Text | 1420 | 1084 |
| LMArena Creative Writing | 1401 | 1047 |
| LMArena Multi-Turn | 1408 | 1050 |
| EQ-Bench Creative Writing | 1436 | — |
Frequently asked questions
Is DeepSeek-V3.1 better than Llama 2-13B?
DeepSeek-V3.1 is the stronger model overall, scoring 42.8 to 29.6 on the Noometry Index.
Is DeepSeek-V3.1 or Llama 2-13B better for coding?
DeepSeek-V3.1 scores higher on coding benchmarks: 40.3 versus 30.9 in the Noometry coding category.
How many benchmarks do DeepSeek-V3.1 and Llama 2-13B share?
19 benchmarks have published results for both models. DeepSeek-V3.1 has 27 scored results on Noometry and Llama 2-13B has 32.