Model Comparison

Comprehensive side-by-side analysis of model capabilities and performance

DeepSeek R1 Zero

DeepSeek

DeepSeek R1 Zero is a language model developed by DeepSeek. It achieves strong performance with an average score of 76.5% across 4 benchmarks. It excels particularly in MATH-500 (95.9%), AIME 2024 (86.7%), GPQA (73.3%). It's licensed for commercial use, making it suitable for enterprise applications. Released in 2025, it represents DeepSeek's latest advancement in AI technology.

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.

DeepSeek R1 Zero

DeepSeek

2025-01-20

Gemma 3 1B

Google

2025-03-12

1 month newer

Average performance across 2 common benchmarks

DeepSeek R1 Zero

+51.1%

Average Score:61.7%

Gemma 3 1B

Average Score:10.5%

Performance comparison across key benchmark categories

DeepSeek R1 Zero

math

+53.7%

95.9%

general

+62.2%

80.0%

code

+7.0%

50.0%

Gemma 3 1B

math

42.2%

general

17.8%

code

43.0%

Benchmark Scores - Detailed View

Side-by-side comparison of all benchmark scores

Provider Availability & Performance

Available providers and their performance metrics

DeepSeek R1 Zero

0 providers

Gemma 3 1B

0 providers

DeepSeek R1 Zero

+51.1%

Avg Score:61.7%

Providers:0

Gemma 3 1B

Avg Score:10.5%

Providers:0