Model comparison
Llama 2-34B vs Llama 3-8B
Llama 3-8B has enough public results to be ranked (#338); Llama 2-34B does not yet, so treat this comparison as directional.
Last verified . 6 shared benchmarks.
MetaSummary
- They share 6 benchmarks with published results for both.
- Llama 3-8B has downloadable open weights; the other is API-only.
Side by side
| Llama 2-34B | Llama 3-8B | |
|---|---|---|
| Provider | Meta | Meta |
| Noometry Index | — | 25.3 |
| Released | 2023-07-18 | 2024-04-18 |
| Weights | Proprietary | Open |
| Context window | — | 8K |
| Max output | — | 8K |
| Input $ / M tokens | — | $0.14 |
| Output $ / M tokens | — | $0.14 |
| Results tracked | 10 | 34 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Not comparable
Llama 2-34B: —, Llama 3-8B: 30.5 (#285)
| Benchmark | Llama 2-34B | Llama 3-8B |
|---|---|---|
| BigCodeBench Instruct | — | 31.9% |
| LMArena Coding | — | 1152 |
| BigCodeBench Complete | — | 36.9% |
| HumanEval+ | — | 56.7% |
| MBPP+ | — | 54.8% |
Reasoning Not comparable
Llama 2-34B: —, Llama 3-8B: 13.8 (#321)
| Benchmark | Llama 2-34B | Llama 3-8B |
|---|---|---|
| Epoch Capabilities Index | 104.81 | 116.33 |
| WinoGrande | 76.7% | 75.7% |
| Chess Puzzles | — | 0% |
| LMArena Hard Prompts | — | 1133 |
| DTBench | — | 43.9% |
| Adversarial NLI | — | 57.3% |
| BIG-Bench Hard | 44.1% | — |
| ForecastBench | — | 58.6 |
| PIQA | 81.9% | — |
Math Not comparable
Llama 2-34B: —, Llama 3-8B: 8.3 (#317)
| Benchmark | Llama 2-34B | Llama 3-8B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | — | 1.9% |
| LMArena Math | — | 1151 |
| MATH Level 5 | — | 6.1% |
| GSM8K | 42.2% | — |
Knowledge Not comparable
Llama 2-34B: —, Llama 3-8B: 7.8 (#303)
| Benchmark | Llama 2-34B | Llama 3-8B |
|---|---|---|
| ARC (AI2) Challenge | 54.5% | 82.8% |
| MMLU | 62.6% | 68.8% |
| OpenBookQA | 58.2% | 82.6% |
| TriviaQA | 84.6% | 67.7% |
| GPQA Diamond | — | 26.1% |
| LMArena Expert | — | 1113 |
| BoolQ | 83.7% | — |
Multilingual Not comparable
Llama 2-34B: —, Llama 3-8B: 31.0 (#256)
| Benchmark | Llama 2-34B | Llama 3-8B |
|---|---|---|
| LMArena Non-English | — | 1098 |
| LMArena Chinese | — | 1076 |
| LMArena French | — | 1158 |
| LMArena German | — | 1104 |
| LMArena Japanese | — | 967 |
| LMArena Korean | — | 1003 |
| LMArena Russian | — | 1109 |
| LMArena Spanish | — | 1172 |
Instruction Following Not comparable
Llama 2-34B: —, Llama 3-8B: 58.7 (#255)
| Benchmark | Llama 2-34B | Llama 3-8B |
|---|---|---|
| LMArena Instruction Following | — | 1127 |
Long Context Not comparable
Llama 2-34B: —, Llama 3-8B: 34.2 (#246)
| Benchmark | Llama 2-34B | Llama 3-8B |
|---|---|---|
| LMArena Longer Query | — | 1128 |
Writing & Preference Not comparable
Llama 2-34B: —, Llama 3-8B: 37.7 (#251)
| Benchmark | Llama 2-34B | Llama 3-8B |
|---|---|---|
| LMArena Text | — | 1166 |
| LMArena Creative Writing | — | 1150 |
| LMArena Multi-Turn | — | 1152 |
Frequently asked questions
Is Llama 2-34B better than Llama 3-8B?
Llama 3-8B has enough public results to be ranked (#338); Llama 2-34B does not yet, so treat this comparison as directional.
How many benchmarks do Llama 2-34B and Llama 3-8B share?
6 benchmarks have published results for both models. Llama 2-34B has 10 scored results on Noometry and Llama 3-8B has 34.