Model comparison
Llama 13b vs Llama 2-7B
Llama 2-7B is the stronger model overall, scoring 28.6 to 23.9 on the Noometry Index.
Last verified . 21 shared benchmarks.
MetaSummary
- They share 21 benchmarks with published results for both. Llama 13b scores higher in 0 categories and Llama 2-7B in 6 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Llama 2-7B leads 28.1 to 13.9.
Side by side
| Llama 13b | Llama 2-7B | |
|---|---|---|
| Provider | Meta | Meta |
| Noometry Index | 23.9 | 28.6 |
| Released | 2023-02-24 | 2023-07-18 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 21 | 29 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Llama 2-7B leads
Llama 13b: 20.9 (#332), Llama 2-7B: 28.7 (#302)
| Benchmark | Llama 13b | Llama 2-7B |
|---|---|---|
| LMArena Coding | 683 | 1002 |
Reasoning Llama 2-7B leads
Llama 13b: 13.4 (#327), Llama 2-7B: 15.1 (#309)
| Benchmark | Llama 13b | Llama 2-7B |
|---|---|---|
| LMArena Hard Prompts | 728 | 1009 |
| BIG-Bench Hard | 37.9% | 39.2% |
| Epoch Capabilities Index | 100.09 | 98.54 |
| HellaSwag | 79.2% | 77.2% |
| LAMBADA | 75.2% | 73.3% |
| PIQA | 80.1% | 78.8% |
| WinoGrande | 73% | 69.2% |
| Chess Puzzles | — | 0% |
Math Llama 2-7B leads
Llama 13b: 24.4 (#259), Llama 2-7B: 28.2 (#231)
| Benchmark | Llama 13b | Llama 2-7B |
|---|---|---|
| LMArena Math | 838 | 1042 |
| GSM8K | 20.6% | 16.7% |
Knowledge Not comparable
Llama 13b: —, Llama 2-7B: 27.7 (#243)
| Benchmark | Llama 13b | Llama 2-7B |
|---|---|---|
| ARC (AI2) Challenge | 52.7% | 45.9% |
| BoolQ | 78.7% | 77.9% |
| MMLU | 47.7% | 45.8% |
| OpenBookQA | 56.4% | 58.6% |
| TriviaQA | 77.9% | 73.7% |
| LMArena Expert | — | 1036 |
Multimodal Not comparable
Llama 13b: —, Llama 2-7B: —
| Benchmark | Llama 13b | Llama 2-7B |
|---|---|---|
| ScienceQA | 43.3% | 43.1% |
Multilingual Llama 2-7B leads
Llama 13b: 16.8 (#292), Llama 2-7B: 23.9 (#288)
| Benchmark | Llama 13b | Llama 2-7B |
|---|---|---|
| LMArena Non-English | 819 | 973 |
| LMArena Chinese | — | 973 |
| LMArena French | — | 969 |
| LMArena German | — | 978 |
| LMArena Russian | — | 995 |
| LMArena Spanish | — | 1007 |
Instruction Following Llama 2-7B leads
Llama 13b: 36.9 (#300), Llama 2-7B: 51.0 (#293)
| Benchmark | Llama 13b | Llama 2-7B |
|---|---|---|
| LMArena Instruction Following | 781 | 1006 |
Long Context Not comparable
Llama 13b: —, Llama 2-7B: 30.4 (#282)
| Benchmark | Llama 13b | Llama 2-7B |
|---|---|---|
| LMArena Longer Query | — | 999 |
Writing & Preference Llama 2-7B leads
Llama 13b: 13.9 (#307), Llama 2-7B: 28.1 (#293)
| Benchmark | Llama 13b | Llama 2-7B |
|---|---|---|
| LMArena Text | 834 | 1053 |
| LMArena Creative Writing | 794 | 1033 |
| LMArena Multi-Turn | 753 | 1029 |
Frequently asked questions
Is Llama 13b better than Llama 2-7B?
Llama 2-7B is the stronger model overall, scoring 28.6 to 23.9 on the Noometry Index.
Is Llama 13b or Llama 2-7B better for coding?
Llama 2-7B scores higher on coding benchmarks: 28.7 versus 20.9 in the Noometry coding category.
How many benchmarks do Llama 13b and Llama 2-7B share?
21 benchmarks have published results for both models. Llama 13b has 21 scored results on Noometry and Llama 2-7B has 29.