Model comparison
DeepSeek-V2.5 (Sep 2024) vs Llama 13b
DeepSeek-V2.5 (Sep 2024) is the stronger model overall, scoring 37.6 to 24.4 on the Noometry Index.
Last verified . 8 shared benchmarks.
DeepSeek
MetaSummary
- They share 8 benchmarks with published results for both. DeepSeek-V2.5 (Sep 2024) scores higher in 6 categories and Llama 13b in 0 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where DeepSeek-V2.5 (Sep 2024) leads 49.8 to 13.8.
Side by side
| DeepSeek-V2.5 (Sep 2024) | Llama 13b | |
|---|---|---|
| Provider | DeepSeek | Meta |
| Noometry Index | 37.6 | 24.4 |
| Released | 2024-09-06 | 2023-02-24 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 22 | 21 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 31.7 (#281), Llama 13b: 21.4 (#337)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Llama 13b |
|---|---|---|
| LMArena Coding | 1309 | 683 |
| Aider Polyglot | 17.8% | — |
| BigCodeBench Instruct | 48.6% | — |
| BigCodeBench Complete | 53.2% | — |
| HumanEval+ | 83.5% | — |
| MBPP+ | 74.1% | — |
Reasoning DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 25.6 (#145), Llama 13b: 14.0 (#329)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Llama 13b |
|---|---|---|
| LMArena Hard Prompts | 1289 | 728 |
| BIG-Bench Hard | — | 37.9% |
| Epoch Capabilities Index | — | 100.58 |
| HellaSwag | — | 79.2% |
| LAMBADA | — | 75.2% |
| PIQA | — | 80.1% |
| WinoGrande | — | 73% |
Math DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 35.9 (#177), Llama 13b: 26.7 (#256)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Llama 13b |
|---|---|---|
| LMArena Math | 1288 | 838 |
| GSM8K | — | 20.6% |
Knowledge Not comparable
DeepSeek-V2.5 (Sep 2024): 34.8 (#193), Llama 13b: —
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Llama 13b |
|---|---|---|
| LMArena Expert | 1266 | — |
| ARC (AI2) Challenge | — | 52.7% |
| BoolQ | — | 78.7% |
| MMLU | — | 47.7% |
| OpenBookQA | — | 56.4% |
| TriviaQA | — | 77.9% |
Multimodal Not comparable
DeepSeek-V2.5 (Sep 2024): —, Llama 13b: —
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Llama 13b |
|---|---|---|
| ScienceQA | — | 43.3% |
Multilingual DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 42.5 (#193), Llama 13b: 16.6 (#297)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Llama 13b |
|---|---|---|
| LMArena Non-English | 1273 | 819 |
| LMArena Chinese | 1318 | — |
| LMArena French | 1289 | — |
| LMArena German | 1258 | — |
| LMArena Japanese | 1228 | — |
| LMArena Korean | 1209 | — |
| LMArena Russian | 1289 | — |
| LMArena Spanish | 1248 | — |
Instruction Following DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 67.5 (#194), Llama 13b: 36.7 (#305)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Llama 13b |
|---|---|---|
| LMArena Instruction Following | 1280 | 781 |
Long Context Not comparable
DeepSeek-V2.5 (Sep 2024): 39.5 (#174), Llama 13b: —
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Llama 13b |
|---|---|---|
| LMArena Longer Query | 1301 | — |
Writing & Preference DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 49.8 (#187), Llama 13b: 13.8 (#312)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Llama 13b |
|---|---|---|
| LMArena Text | 1294 | 834 |
| LMArena Creative Writing | 1285 | 794 |
| LMArena Multi-Turn | 1297 | 753 |
Frequently asked questions
Is DeepSeek-V2.5 (Sep 2024) better than Llama 13b?
DeepSeek-V2.5 (Sep 2024) is the stronger model overall, scoring 37.6 to 24.4 on the Noometry Index.
Is DeepSeek-V2.5 (Sep 2024) or Llama 13b better for coding?
DeepSeek-V2.5 (Sep 2024) scores higher on coding benchmarks: 31.7 versus 21.4 in the Noometry coding category.
How many benchmarks do DeepSeek-V2.5 (Sep 2024) and Llama 13b share?
8 benchmarks have published results for both models. DeepSeek-V2.5 (Sep 2024) has 22 scored results on Noometry and Llama 13b has 21.