Model comparison
DeepSeek-R1 vs Llama 2-13B
DeepSeek-R1 is the stronger model overall, scoring 42.3 to 29.6 on the Noometry Index.
Last verified . 18 shared benchmarks.
DeepSeek
MetaSummary
- They share 18 benchmarks with published results for both. DeepSeek-R1 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-R1 leads 61.4 to 29.8.
- Llama 2-13B has downloadable open weights; the other is API-only.
Side by side
| DeepSeek-R1 | Llama 2-13B | |
|---|---|---|
| Provider | DeepSeek | Meta |
| Noometry Index | 42.3 | 29.6 |
| Released | 2025-01-20 | 2023-07-18 |
| Weights | Proprietary | Open |
| Context window | 164K | — |
| Max output | 64K | — |
| Input $ / M tokens | $0.50 | — |
| Output $ / M tokens | $2.15 | — |
| Results tracked | 52 | 32 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek-R1 leads
DeepSeek-R1: 46.3 (#68), Llama 2-13B: 30.9 (#291)
| Benchmark | DeepSeek-R1 | Llama 2-13B |
|---|---|---|
| LMArena Coding | 1427 | 1062 |
| Aider Polyglot | 71.4% | — |
| SciCode | 35.7% | — |
| WeirdML | 41.6% | — |
| LiveBench Coding | 66.7% | — |
| ALE-Bench | 804.12 | — |
| AlgoTune | 1.7 | — |
Agentic & Tool Use Not comparable
DeepSeek-R1: 30.7 (#75), Llama 2-13B: —
| Benchmark | DeepSeek-R1 | Llama 2-13B |
|---|---|---|
| DeepResearch Bench | 35.1% | — |
| BALROG | 34.9% | — |
| METR Time Horizons | 53.8% | — |
Reasoning DeepSeek-R1 leads
DeepSeek-R1: 18.6 (#278), Llama 2-13B: 12.8 (#337)
| Benchmark | DeepSeek-R1 | Llama 2-13B |
|---|---|---|
| LMArena Hard Prompts | 1416 | 1051 |
| Epoch Capabilities Index | 141.29 | 106.17 |
| ARC-AGI-2 | 1.3% | — |
| SimpleBench | 40.8% | — |
| Kagi LLM Benchmark | 69.4% | — |
| ARC-AGI-1 | 21.2% | — |
| CritPt | 1.1% | — |
| Chess Puzzles | — | 0% |
| LiveBench Reasoning | 83.2% | — |
| DTBench | — | 42.2% |
| LiveBench Data Analysis | 69.8% | — |
| BIG-Bench Hard | — | 58.2% |
| ForecastBench | 60 | — |
| HellaSwag | — | 80.7% |
| LAMBADA | — | 76.5% |
| LiveBench | 71.6% | — |
| PIQA | — | 80.8% |
| WinoGrande | — | 72.8% |
Math DeepSeek-R1 leads
DeepSeek-R1: 43.8 (#79), Llama 2-13B: 31.1 (#229)
| Benchmark | DeepSeek-R1 | Llama 2-13B |
|---|---|---|
| LMArena Math | 1400 | 1065 |
| OTIS Mock AIME 2024-2025 | 66.4% | — |
| Omni-MATH | 42.4% | — |
| LiveBench Math | 80.7% | — |
| MATH Level 5 | 96.6% | — |
| GSM8K | — | 36.9% |
Knowledge DeepSeek-R1 leads
DeepSeek-R1: 44.5 (#87), Llama 2-13B: 28.1 (#249)
| Benchmark | DeepSeek-R1 | Llama 2-13B |
|---|---|---|
| LMArena Expert | 1394 | 1030 |
| GPQA Diamond | 76.3% | — |
| MMLU-Pro | 79.3% | — |
| Confabulations | 12.7% | — |
| Vectara Hallucination Rate | 11.3% | — |
| GPQA (HELM) | 66.6% | — |
| ARC (AI2) Challenge | — | 60.3% |
| BoolQ | — | 82.4% |
| MMLU | — | 55.6% |
| OpenBookQA | — | 57% |
| TriviaQA | — | 79.6% |
Multimodal Not comparable
DeepSeek-R1: —, Llama 2-13B: —
| Benchmark | DeepSeek-R1 | Llama 2-13B |
|---|---|---|
| ScienceQA | — | 55.8% |
Multilingual DeepSeek-R1 leads
DeepSeek-R1: 52.4 (#85), Llama 2-13B: 26.5 (#279)
| Benchmark | DeepSeek-R1 | Llama 2-13B |
|---|---|---|
| LMArena Non-English | 1412 | 1024 |
| LMArena Chinese | 1442 | 1001 |
| LMArena French | 1417 | 1044 |
| LMArena German | 1404 | 1009 |
| LMArena Japanese | 1391 | 894 |
| LMArena Korean | 1360 | 953 |
| LMArena Russian | 1423 | 1055 |
| LMArena Spanish | 1411 | 1087 |
Instruction Following DeepSeek-R1 leads
DeepSeek-R1: 72.0 (#143), Llama 2-13B: 53.3 (#287)
| Benchmark | DeepSeek-R1 | Llama 2-13B |
|---|---|---|
| LMArena Instruction Following | 1382 | 1045 |
| LiveBench Instruction Following | 80.5% | — |
| IFEval | 78.4% | — |
Long Context DeepSeek-R1 leads
DeepSeek-R1: 45.4 (#36), Llama 2-13B: 32.3 (#269)
| Benchmark | DeepSeek-R1 | Llama 2-13B |
|---|---|---|
| LMArena Longer Query | 1391 | 1064 |
| Fiction.LiveBench | 75% | — |
Writing & Preference DeepSeek-R1 leads
DeepSeek-R1: 61.4 (#88), Llama 2-13B: 29.8 (#289)
| Benchmark | DeepSeek-R1 | Llama 2-13B |
|---|---|---|
| LMArena Text | 1428 | 1084 |
| LMArena Creative Writing | 1405 | 1047 |
| LMArena Multi-Turn | 1405 | 1050 |
| Short-Story Creative Writing | 83% | — |
| EQ-Bench Creative Writing | 1500 | — |
| WildBench | 82.8% | — |
| LiveBench Language | 48.5% | — |
Frequently asked questions
Is DeepSeek-R1 better than Llama 2-13B?
DeepSeek-R1 is the stronger model overall, scoring 42.3 to 29.6 on the Noometry Index.
Is DeepSeek-R1 or Llama 2-13B better for coding?
DeepSeek-R1 scores higher on coding benchmarks: 46.3 versus 30.9 in the Noometry coding category.
How many benchmarks do DeepSeek-R1 and Llama 2-13B share?
18 benchmarks have published results for both models. DeepSeek-R1 has 52 scored results on Noometry and Llama 2-13B has 32.