Model comparison
GPT-3.5-turbo vs Phi-2
GPT-3.5-turbo has enough public results to be ranked (#350); Phi-2 does not yet, so treat this comparison as directional.
Last verified . 10 shared benchmarks.
MicrosoftSummary
- They share 10 benchmarks with published results for both.
- Phi-2 has downloadable open weights; the other is API-only.
Side by side
| GPT-3.5-turbo | Phi-2 | |
|---|---|---|
| Provider | OpenAI | Microsoft |
| Noometry Index | 23.2 | — |
| Released | 2023-03-01 | 2023-12-12 |
| Weights | Proprietary | Open |
| Context window | 16K | — |
| Max output | 4K | — |
| Input $ / M tokens | $0.50 | — |
| Output $ / M tokens | $1.50 | — |
| Results tracked | 44 | 11 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Not comparable
GPT-3.5-turbo: 23.9 (#331), Phi-2: —
| Benchmark | GPT-3.5-turbo | Phi-2 |
|---|---|---|
| HumanEval+ | 70.7% | 45.1% |
| MBPP+ | 69.7% | 54.2% |
| WeirdML | 3.5% | — |
| BigCodeBench Instruct | 39.1% | — |
| LMArena Coding | 1136 | — |
| BigCodeBench Complete | 50.6% | — |
Agentic & Tool Use Not comparable
GPT-3.5-turbo: —, Phi-2: —
| Benchmark | GPT-3.5-turbo | Phi-2 |
|---|---|---|
| METR Time Horizons | 21.5% | — |
Reasoning Not comparable
GPT-3.5-turbo: 13.8 (#332), Phi-2: —
| Benchmark | GPT-3.5-turbo | Phi-2 |
|---|---|---|
| Adversarial NLI | 58.1% | 42.5% |
| BIG-Bench Hard | 61.6% | 59.4% |
| Epoch Capabilities Index | 118.55 | 107.94 |
| WinoGrande | 81.6% | 54.7% |
| Chess Puzzles | 0% | — |
| LMArena Hard Prompts | 1108 | — |
| Mystery Game Puzzles | 3% | — |
| DTBench | 48.5% | — |
| LMCA | 9.7% | — |
| CommonsenseQA 2.0 | 57% | — |
| ForecastBench | 50.4 | — |
| HellaSwag | — | 53.6% |
Math Not comparable
GPT-3.5-turbo: 6.3 (#327), Phi-2: —
| Benchmark | GPT-3.5-turbo | Phi-2 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 0% | — |
| OTIS Mock AIME 2024-2025 | 2.2% | — |
| LMArena Math | 1142 | — |
| MATH Level 5 | 15.9% | — |
| GSM8K | 57.8% | — |
Knowledge Not comparable
GPT-3.5-turbo: 10.0 (#303), Phi-2: —
| Benchmark | GPT-3.5-turbo | Phi-2 |
|---|---|---|
| ARC (AI2) Challenge | 87.4% | 75.9% |
| MMLU | 71.4% | 58.4% |
| OpenBookQA | 86% | 73.6% |
| TriviaQA | 85.8% | 45.2% |
| GPQA Diamond | 28% | — |
| LMArena Expert | 1070 | — |
| BoolQ | 87% | — |
Multilingual Not comparable
GPT-3.5-turbo: 31.5 (#258), Phi-2: —
| Benchmark | GPT-3.5-turbo | Phi-2 |
|---|---|---|
| LMArena Non-English | 1108 | — |
| LMArena Chinese | 1075 | — |
| LMArena French | 1118 | — |
| LMArena German | 1090 | — |
| LMArena Japanese | 1043 | — |
| LMArena Korean | 1019 | — |
| LMArena Russian | 1123 | — |
| LMArena Spanish | 1121 | — |
Instruction Following Not comparable
GPT-3.5-turbo: 57.9 (#262), Phi-2: —
| Benchmark | GPT-3.5-turbo | Phi-2 |
|---|---|---|
| LMArena Instruction Following | 1119 | — |
Long Context Not comparable
GPT-3.5-turbo: 34.0 (#254), Phi-2: —
| Benchmark | GPT-3.5-turbo | Phi-2 |
|---|---|---|
| LMArena Longer Query | 1121 | — |
Writing & Preference Not comparable
GPT-3.5-turbo: 25.3 (#305), Phi-2: —
| Benchmark | GPT-3.5-turbo | Phi-2 |
|---|---|---|
| LMArena Text | 1125 | — |
| LMArena Creative Writing | 1092 | — |
| EQ-Bench Creative Writing | 451 | — |
| LMArena Multi-Turn | 1117 | — |
Frequently asked questions
Is GPT-3.5-turbo better than Phi-2?
GPT-3.5-turbo has enough public results to be ranked (#350); Phi-2 does not yet, so treat this comparison as directional.
How many benchmarks do GPT-3.5-turbo and Phi-2 share?
10 benchmarks have published results for both models. GPT-3.5-turbo has 44 scored results on Noometry and Phi-2 has 11.