Model comparison
Gemini 1.0 Pro vs Phi-2
Gemini 1.0 Pro has enough public results to be ranked (#332); Phi-2 does not yet, so treat this comparison as directional.
Last verified . 4 shared benchmarks.
Google
MicrosoftSummary
- They share 4 benchmarks with published results for both.
- Phi-2 has downloadable open weights; the other is API-only.
Side by side
| Gemini 1.0 Pro | Phi-2 | |
|---|---|---|
| Provider | Microsoft | |
| Noometry Index | 27.3 | — |
| Released | 2023-12-13 | 2023-12-12 |
| Weights | Proprietary | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 24 | 11 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Not comparable
Gemini 1.0 Pro: 32.2 (#275), Phi-2: —
| Benchmark | Gemini 1.0 Pro | Phi-2 |
|---|---|---|
| HumanEval+ | 55.5% | 45.1% |
| MBPP+ | 61.4% | 54.2% |
| LMArena Coding | 1108 | — |
Reasoning Not comparable
Gemini 1.0 Pro: 17.1 (#296), Phi-2: —
| Benchmark | Gemini 1.0 Pro | Phi-2 |
|---|---|---|
| Epoch Capabilities Index | 117.04 | 107.94 |
| LMArena Hard Prompts | 1109 | — |
| DTBench | 45.9% | — |
| Adversarial NLI | — | 42.5% |
| BIG-Bench Hard | — | 59.4% |
| HellaSwag | — | 53.6% |
| WinoGrande | — | 54.7% |
Math Not comparable
Gemini 1.0 Pro: 9.3 (#321), Phi-2: —
| Benchmark | Gemini 1.0 Pro | Phi-2 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 1.1% | — |
| LMArena Math | 1132 | — |
| MATH Level 5 | 11.2% | — |
Knowledge Not comparable
Gemini 1.0 Pro: 15.6 (#291), Phi-2: —
| Benchmark | Gemini 1.0 Pro | Phi-2 |
|---|---|---|
| MMLU | 70% | 58.4% |
| GPQA Diamond | 34% | — |
| LMArena Expert | 1059 | — |
| ARC (AI2) Challenge | — | 75.9% |
| OpenBookQA | — | 73.6% |
| TriviaQA | — | 45.2% |
Multilingual Not comparable
Gemini 1.0 Pro: 33.4 (#252), Phi-2: —
| Benchmark | Gemini 1.0 Pro | Phi-2 |
|---|---|---|
| LMArena Non-English | 1138 | — |
| LMArena Chinese | 1124 | — |
| LMArena French | 1145 | — |
| LMArena German | 1125 | — |
| LMArena Japanese | 1023 | — |
| LMArena Russian | 1186 | — |
| LMArena Spanish | 1119 | — |
Instruction Following Not comparable
Gemini 1.0 Pro: 57.6 (#267), Phi-2: —
| Benchmark | Gemini 1.0 Pro | Phi-2 |
|---|---|---|
| LMArena Instruction Following | 1114 | — |
Long Context Not comparable
Gemini 1.0 Pro: 34.3 (#249), Phi-2: —
| Benchmark | Gemini 1.0 Pro | Phi-2 |
|---|---|---|
| LMArena Longer Query | 1132 | — |
Writing & Preference Not comparable
Gemini 1.0 Pro: 36.0 (#264), Phi-2: —
| Benchmark | Gemini 1.0 Pro | Phi-2 |
|---|---|---|
| LMArena Text | 1149 | — |
| LMArena Creative Writing | 1131 | — |
| LMArena Multi-Turn | 1139 | — |
Frequently asked questions
Is Gemini 1.0 Pro better than Phi-2?
Gemini 1.0 Pro has enough public results to be ranked (#332); Phi-2 does not yet, so treat this comparison as directional.
How many benchmarks do Gemini 1.0 Pro and Phi-2 share?
4 benchmarks have published results for both models. Gemini 1.0 Pro has 24 scored results on Noometry and Phi-2 has 11.