Model comparison
DeepSeek-V2.5 (Sep 2024) vs Llama 2-13B
DeepSeek-V2.5 (Sep 2024) is the stronger model overall, scoring 37.6 to 29.6 on the Noometry Index.
Last verified . 17 shared benchmarks.
DeepSeek
MetaSummary
- They share 17 benchmarks with published results for both. DeepSeek-V2.5 (Sep 2024) scores higher in 8 categories and Llama 2-13B in 0 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where DeepSeek-V2.5 (Sep 2024) leads 49.8 to 29.8.
Side by side
| DeepSeek-V2.5 (Sep 2024) | Llama 2-13B | |
|---|---|---|
| Provider | DeepSeek | Meta |
| Noometry Index | 37.6 | 29.6 |
| Released | 2024-09-06 | 2023-07-18 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 22 | 32 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
DeepSeek-V2.5 (Sep 2024): 31.7 (#281), Llama 2-13B: 30.9 (#291)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Llama 2-13B |
|---|---|---|
| LMArena Coding | 1309 | 1062 |
| 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 2-13B: 12.8 (#337)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Llama 2-13B |
|---|---|---|
| LMArena Hard Prompts | 1289 | 1051 |
| Chess Puzzles | — | 0% |
| DTBench | — | 42.2% |
| BIG-Bench Hard | — | 58.2% |
| Epoch Capabilities Index | — | 106.17 |
| HellaSwag | — | 80.7% |
| LAMBADA | — | 76.5% |
| PIQA | — | 80.8% |
| WinoGrande | — | 72.8% |
Math DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 35.9 (#177), Llama 2-13B: 31.1 (#229)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Llama 2-13B |
|---|---|---|
| LMArena Math | 1288 | 1065 |
| GSM8K | — | 36.9% |
Knowledge DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 34.8 (#193), Llama 2-13B: 28.1 (#249)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Llama 2-13B |
|---|---|---|
| LMArena Expert | 1266 | 1030 |
| ARC (AI2) Challenge | — | 60.3% |
| BoolQ | — | 82.4% |
| MMLU | — | 55.6% |
| OpenBookQA | — | 57% |
| TriviaQA | — | 79.6% |
Multimodal Not comparable
DeepSeek-V2.5 (Sep 2024): —, Llama 2-13B: —
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Llama 2-13B |
|---|---|---|
| ScienceQA | — | 55.8% |
Multilingual DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 42.5 (#193), Llama 2-13B: 26.5 (#279)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Llama 2-13B |
|---|---|---|
| LMArena Non-English | 1273 | 1024 |
| LMArena Chinese | 1318 | 1001 |
| LMArena French | 1289 | 1044 |
| LMArena German | 1258 | 1009 |
| LMArena Japanese | 1228 | 894 |
| LMArena Korean | 1209 | 953 |
| LMArena Russian | 1289 | 1055 |
| LMArena Spanish | 1248 | 1087 |
Instruction Following DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 67.5 (#194), Llama 2-13B: 53.3 (#287)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Llama 2-13B |
|---|---|---|
| LMArena Instruction Following | 1280 | 1045 |
Long Context DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 39.5 (#174), Llama 2-13B: 32.3 (#269)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Llama 2-13B |
|---|---|---|
| LMArena Longer Query | 1301 | 1064 |
Writing & Preference DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 49.8 (#187), Llama 2-13B: 29.8 (#289)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Llama 2-13B |
|---|---|---|
| LMArena Text | 1294 | 1084 |
| LMArena Creative Writing | 1285 | 1047 |
| LMArena Multi-Turn | 1297 | 1050 |
Frequently asked questions
Is DeepSeek-V2.5 (Sep 2024) better than Llama 2-13B?
DeepSeek-V2.5 (Sep 2024) is the stronger model overall, scoring 37.6 to 29.6 on the Noometry Index.
Is DeepSeek-V2.5 (Sep 2024) or Llama 2-13B better for coding?
They score almost the same on coding (31.7 vs 30.9); test both on your own repository before choosing.
How many benchmarks do DeepSeek-V2.5 (Sep 2024) and Llama 2-13B share?
17 benchmarks have published results for both models. DeepSeek-V2.5 (Sep 2024) has 22 scored results on Noometry and Llama 2-13B has 32.