Model comparison
Llama 2-34B vs Phi-2
Neither Llama 2-34B nor Phi-2 has enough public benchmark results to be ranked yet; the rows below show what has been published.
Last verified . 7 shared benchmarks.
Meta
MicrosoftSummary
- They share 7 benchmarks with published results for both.
- Phi-2 has downloadable open weights; the other is API-only.
Side by side
| Llama 2-34B | Phi-2 | |
|---|---|---|
| Provider | Meta | Microsoft |
| Noometry Index | — | — |
| Released | 2023-07-18 | 2023-12-12 |
| Weights | Proprietary | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 10 | 11 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Not comparable
Llama 2-34B: —, Phi-2: —
| Benchmark | Llama 2-34B | Phi-2 |
|---|---|---|
| HumanEval+ | — | 45.1% |
| MBPP+ | — | 54.2% |
Reasoning Not comparable
Llama 2-34B: —, Phi-2: —
| Benchmark | Llama 2-34B | Phi-2 |
|---|---|---|
| BIG-Bench Hard | 44.1% | 59.4% |
| Epoch Capabilities Index | 104.81 | 107.62 |
| WinoGrande | 76.7% | 54.7% |
| Adversarial NLI | — | 42.5% |
| HellaSwag | — | 53.6% |
| PIQA | 81.9% | — |
Math Not comparable
Llama 2-34B: —, Phi-2: —
| Benchmark | Llama 2-34B | Phi-2 |
|---|---|---|
| GSM8K | 42.2% | — |
Knowledge Not comparable
Llama 2-34B: —, Phi-2: —
| Benchmark | Llama 2-34B | Phi-2 |
|---|---|---|
| ARC (AI2) Challenge | 54.5% | 75.9% |
| MMLU | 62.6% | 58.4% |
| OpenBookQA | 58.2% | 73.6% |
| TriviaQA | 84.6% | 45.2% |
| BoolQ | 83.7% | — |
Frequently asked questions
Is Llama 2-34B better than Phi-2?
Neither Llama 2-34B nor Phi-2 has enough public benchmark results to be ranked yet; the rows below show what has been published.
How many benchmarks do Llama 2-34B and Phi-2 share?
7 benchmarks have published results for both models. Llama 2-34B has 10 scored results on Noometry and Phi-2 has 11.