Model comparison
Llama 13b vs Llama 2-13B
Llama 2-13B is the stronger model overall, scoring 29.6 to 24.4 on the Noometry Index.
Last verified . 21 shared benchmarks.
MetaSummary
- They share 21 benchmarks with published results for both. Llama 13b scores higher in 1 category and Llama 2-13B in 5 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in instruction following, where Llama 2-13B leads 53.3 to 36.7.
Side by side
| Llama 13b | Llama 2-13B | |
|---|---|---|
| Provider | Meta | Meta |
| Noometry Index | 24.4 | 29.6 |
| Released | 2023-02-24 | 2023-07-18 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 21 | 32 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Llama 2-13B leads
Llama 13b: 21.4 (#337), Llama 2-13B: 30.9 (#291)
| Benchmark | Llama 13b | Llama 2-13B |
|---|---|---|
| LMArena Coding | 683 | 1062 |
Reasoning Llama 13b leads
Llama 13b: 14.0 (#329), Llama 2-13B: 12.8 (#337)
| Benchmark | Llama 13b | Llama 2-13B |
|---|---|---|
| LMArena Hard Prompts | 728 | 1051 |
| BIG-Bench Hard | 37.9% | 58.2% |
| Epoch Capabilities Index | 100.58 | 106.17 |
| HellaSwag | 79.2% | 80.7% |
| LAMBADA | 75.2% | 76.5% |
| PIQA | 80.1% | 80.8% |
| WinoGrande | 73% | 72.8% |
| Chess Puzzles | — | 0% |
| DTBench | — | 42.2% |
Math Llama 2-13B leads
Llama 13b: 26.7 (#256), Llama 2-13B: 31.1 (#229)
| Benchmark | Llama 13b | Llama 2-13B |
|---|---|---|
| LMArena Math | 838 | 1065 |
| GSM8K | 20.6% | 36.9% |
Knowledge Not comparable
Llama 13b: —, Llama 2-13B: 28.1 (#249)
| Benchmark | Llama 13b | Llama 2-13B |
|---|---|---|
| ARC (AI2) Challenge | 52.7% | 60.3% |
| BoolQ | 78.7% | 82.4% |
| MMLU | 47.7% | 55.6% |
| OpenBookQA | 56.4% | 57% |
| TriviaQA | 77.9% | 79.6% |
| LMArena Expert | — | 1030 |
Multimodal Not comparable
Llama 13b: —, Llama 2-13B: —
| Benchmark | Llama 13b | Llama 2-13B |
|---|---|---|
| ScienceQA | 43.3% | 55.8% |
Multilingual Llama 2-13B leads
Llama 13b: 16.6 (#297), Llama 2-13B: 26.5 (#279)
| Benchmark | Llama 13b | Llama 2-13B |
|---|---|---|
| LMArena Non-English | 819 | 1024 |
| LMArena Chinese | — | 1001 |
| LMArena French | — | 1044 |
| LMArena German | — | 1009 |
| LMArena Japanese | — | 894 |
| LMArena Korean | — | 953 |
| LMArena Russian | — | 1055 |
| LMArena Spanish | — | 1087 |
Instruction Following Llama 2-13B leads
Llama 13b: 36.7 (#305), Llama 2-13B: 53.3 (#287)
| Benchmark | Llama 13b | Llama 2-13B |
|---|---|---|
| LMArena Instruction Following | 781 | 1045 |
Long Context Not comparable
Llama 13b: —, Llama 2-13B: 32.3 (#269)
| Benchmark | Llama 13b | Llama 2-13B |
|---|---|---|
| LMArena Longer Query | — | 1064 |
Writing & Preference Llama 2-13B leads
Llama 13b: 13.8 (#312), Llama 2-13B: 29.8 (#289)
| Benchmark | Llama 13b | Llama 2-13B |
|---|---|---|
| LMArena Text | 834 | 1084 |
| LMArena Creative Writing | 794 | 1047 |
| LMArena Multi-Turn | 753 | 1050 |
Frequently asked questions
Is Llama 13b better than Llama 2-13B?
Llama 2-13B is the stronger model overall, scoring 29.6 to 24.4 on the Noometry Index.
Is Llama 13b or Llama 2-13B better for coding?
Llama 2-13B scores higher on coding benchmarks: 30.9 versus 21.4 in the Noometry coding category.
How many benchmarks do Llama 13b and Llama 2-13B share?
21 benchmarks have published results for both models. Llama 13b has 21 scored results on Noometry and Llama 2-13B has 32.