Model comparison
Llama 13b vs Llama 3-8B
Llama 3-8B is the stronger model overall, scoring 25.5 to 24.4 on the Noometry Index.
Last verified . 14 shared benchmarks.
MetaSummary
- They share 14 benchmarks with published results for both. Llama 13b scores higher in 1 category and Llama 3-8B in 5 categories; 5 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Llama 3-8B leads 37.5 to 13.8.
Side by side
| Llama 13b | Llama 3-8B | |
|---|---|---|
| Provider | Meta | Meta |
| Noometry Index | 24.4 | 25.5 |
| Released | 2023-02-24 | 2024-04-18 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 21 | 34 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Llama 3-8B leads
Llama 13b: 21.4 (#337), Llama 3-8B: 31.0 (#289)
| Benchmark | Llama 13b | Llama 3-8B |
|---|---|---|
| LMArena Coding | 683 | 1152 |
| BigCodeBench Instruct | — | 31.9% |
| BigCodeBench Complete | — | 36.9% |
| HumanEval+ | — | 56.7% |
| MBPP+ | — | 54.8% |
Reasoning Too close to call
Llama 13b: 14.0 (#329), Llama 3-8B: 14.3 (#326)
| Benchmark | Llama 13b | Llama 3-8B |
|---|---|---|
| LMArena Hard Prompts | 728 | 1133 |
| Epoch Capabilities Index | 100.58 | 116.45 |
| WinoGrande | 73% | 75.7% |
| Chess Puzzles | — | 0% |
| DTBench | — | 43.9% |
| Adversarial NLI | — | 57.3% |
| BIG-Bench Hard | 37.9% | — |
| ForecastBench | — | 58.6 |
| HellaSwag | 79.2% | — |
| LAMBADA | 75.2% | — |
| PIQA | 80.1% | — |
Math Llama 13b leads
Llama 13b: 26.7 (#256), Llama 3-8B: 8.8 (#323)
| Benchmark | Llama 13b | Llama 3-8B |
|---|---|---|
| LMArena Math | 838 | 1151 |
| OTIS Mock AIME 2024-2025 | — | 1.9% |
| MATH Level 5 | — | 6.1% |
| GSM8K | 20.6% | — |
Knowledge Not comparable
Llama 13b: —, Llama 3-8B: 7.8 (#308)
| Benchmark | Llama 13b | Llama 3-8B |
|---|---|---|
| ARC (AI2) Challenge | 52.7% | 82.8% |
| MMLU | 47.7% | 68.8% |
| OpenBookQA | 56.4% | 82.6% |
| TriviaQA | 77.9% | 67.7% |
| GPQA Diamond | — | 26.1% |
| LMArena Expert | — | 1113 |
| BoolQ | 78.7% | — |
Multimodal Not comparable
Llama 13b: —, Llama 3-8B: —
| Benchmark | Llama 13b | Llama 3-8B |
|---|---|---|
| ScienceQA | 43.3% | — |
Multilingual Llama 3-8B leads
Llama 13b: 16.6 (#297), Llama 3-8B: 30.8 (#261)
| Benchmark | Llama 13b | Llama 3-8B |
|---|---|---|
| LMArena Non-English | 819 | 1098 |
| LMArena Chinese | — | 1076 |
| LMArena French | — | 1159 |
| LMArena German | — | 1104 |
| LMArena Japanese | — | 967 |
| LMArena Korean | — | 1004 |
| LMArena Russian | — | 1109 |
| LMArena Spanish | — | 1173 |
Instruction Following Llama 3-8B leads
Llama 13b: 36.7 (#305), Llama 3-8B: 58.4 (#260)
| Benchmark | Llama 13b | Llama 3-8B |
|---|---|---|
| LMArena Instruction Following | 781 | 1127 |
Long Context Not comparable
Llama 13b: —, Llama 3-8B: 34.2 (#251)
| Benchmark | Llama 13b | Llama 3-8B |
|---|---|---|
| LMArena Longer Query | — | 1128 |
Writing & Preference Llama 3-8B leads
Llama 13b: 13.8 (#312), Llama 3-8B: 37.5 (#256)
| Benchmark | Llama 13b | Llama 3-8B |
|---|---|---|
| LMArena Text | 834 | 1166 |
| LMArena Creative Writing | 794 | 1150 |
| LMArena Multi-Turn | 753 | 1152 |
Frequently asked questions
Is Llama 13b better than Llama 3-8B?
Llama 3-8B is the stronger model overall, scoring 25.5 to 24.4 on the Noometry Index.
Is Llama 13b or Llama 3-8B better for coding?
Llama 3-8B scores higher on coding benchmarks: 31.0 versus 21.4 in the Noometry coding category.
How many benchmarks do Llama 13b and Llama 3-8B share?
14 benchmarks have published results for both models. Llama 13b has 21 scored results on Noometry and Llama 3-8B has 34.