Model Comparison

Gemma 3 1B

Google

Gemma 3 1B is a language model developed by Google. The model shows competitive results across 18 benchmarks. It excels particularly in IFEval (80.2%), GSM8k (62.8%), Natural2Code (56.0%). It's licensed for commercial use, making it suitable for enterprise applications. Released in 2025, it represents Google's latest advancement in AI technology.

Phi 4 Mini

Microsoft

Phi 4 Mini is a language model developed by Microsoft. It achieves strong performance with an average score of 65.4% across 17 benchmarks. It excels particularly in GSM8k (88.6%), ARC-C (83.7%), BoolQ (81.2%). It's licensed for commercial use, making it suitable for enterprise applications. Released in 2025, it represents Microsoft's latest advancement in AI technology.