Google AI Models

Complete comparison of all 4 AI models from Google, with average MMLU score of 43.3% and performance ranging from 0% to 87% general knowledge.

4 Models
Avg MMLU: 43.3%
Best: Bard Pro

About Google AI Models

Google offers 4 AI models ranging from compact, efficient options to flagship models with cutting-edge capabilities. Their model lineup provides solutions for every use case, with performance ranging from 0% to 87% on MMLU benchmarks.

The top-performing model, Bard Pro, achieves 85.5% on coding benchmarks and 87.2% on reasoning tasks, making it competitive with other industry leaders. For users seeking balanced performance, Imagen offers excellent capabilities with a 0% MMLU score.