Model comparison
Claude 3 Sonnet vs Phi-2
Claude 3 Sonnet has enough public results to be ranked (#319); Phi-2 does not yet, so treat this comparison as directional.
Last verified . 5 shared benchmarks.
MicrosoftSummary
- They share 5 benchmarks with published results for both.
- Phi-2 has downloadable open weights; the other is API-only.
Side by side
| Claude 3 Sonnet | Phi-2 | |
|---|---|---|
| Provider | Anthropic | Microsoft |
| Noometry Index | 29.0 | — |
| Released | 2024-02-29 | 2023-12-12 |
| Weights | Proprietary | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 30 | 11 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Not comparable
Claude 3 Sonnet: 29.6 (#302), Phi-2: —
| Benchmark | Claude 3 Sonnet | Phi-2 |
|---|---|---|
| HumanEval+ | 64% | 45.1% |
| MBPP+ | 69.3% | 54.2% |
| WeirdML | 10.2% | — |
| BigCodeBench Instruct | 42.7% | — |
| LMArena Coding | 1223 | — |
| BigCodeBench Complete | 53.8% | — |
Reasoning Not comparable
Claude 3 Sonnet: 20.5 (#237), Phi-2: —
| Benchmark | Claude 3 Sonnet | Phi-2 |
|---|---|---|
| Epoch Capabilities Index | 120.7 | 107.94 |
| WinoGrande | 75.1% | 54.7% |
| LMArena Hard Prompts | 1197 | — |
| DTBench | 53.6% | — |
| Adversarial NLI | — | 42.5% |
| BIG-Bench Hard | — | 59.4% |
| HellaSwag | — | 53.6% |
Math Not comparable
Claude 3 Sonnet: 10.7 (#310), Phi-2: —
| Benchmark | Claude 3 Sonnet | Phi-2 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 2.5% | — |
| LMArena Math | 1213 | — |
| MATH Level 5 | 18.2% | — |
Knowledge Not comparable
Claude 3 Sonnet: 21.1 (#276), Phi-2: —
| Benchmark | Claude 3 Sonnet | Phi-2 |
|---|---|---|
| MMLU | 75.9% | 58.4% |
| GPQA Diamond | 40.6% | — |
| LMArena Expert | 1173 | — |
| ARC (AI2) Challenge | — | 75.9% |
| OpenBookQA | — | 73.6% |
| TriviaQA | — | 45.2% |
Multimodal Not comparable
Claude 3 Sonnet: 25.2 (#125), Phi-2: —
| Benchmark | Claude 3 Sonnet | Phi-2 |
|---|---|---|
| LMArena Vision | 984 | — |
Multilingual Not comparable
Claude 3 Sonnet: 37.8 (#234), Phi-2: —
| Benchmark | Claude 3 Sonnet | Phi-2 |
|---|---|---|
| LMArena Non-English | 1205 | — |
| LMArena Chinese | 1189 | — |
| LMArena French | 1229 | — |
| LMArena German | 1204 | — |
| LMArena Japanese | 1131 | — |
| LMArena Korean | 1128 | — |
| LMArena Russian | 1227 | — |
| LMArena Spanish | 1204 | — |
Instruction Following Not comparable
Claude 3 Sonnet: 62.8 (#235), Phi-2: —
| Benchmark | Claude 3 Sonnet | Phi-2 |
|---|---|---|
| LMArena Instruction Following | 1199 | — |
Long Context Not comparable
Claude 3 Sonnet: 36.7 (#228), Phi-2: —
| Benchmark | Claude 3 Sonnet | Phi-2 |
|---|---|---|
| LMArena Longer Query | 1211 | — |
Writing & Preference Not comparable
Claude 3 Sonnet: 42.1 (#238), Phi-2: —
| Benchmark | Claude 3 Sonnet | Phi-2 |
|---|---|---|
| LMArena Text | 1218 | — |
| LMArena Creative Writing | 1186 | — |
| LMArena Multi-Turn | 1227 | — |
Frequently asked questions
Is Claude 3 Sonnet better than Phi-2?
Claude 3 Sonnet has enough public results to be ranked (#319); Phi-2 does not yet, so treat this comparison as directional.
How many benchmarks do Claude 3 Sonnet and Phi-2 share?
5 benchmarks have published results for both models. Claude 3 Sonnet has 30 scored results on Noometry and Phi-2 has 11.