Google: Gemma 4 31B

Gemma 4 31B from Google is a versatile AI model capable of handling text, images, and video inputs, with a substantial context length of 262,144 tokens. It supports both reasoning and tool integration, making it suitable for complex tasks that require understanding across multiple modalities. This model excels in scenarios where you need a robust AI with broad capabilities, but at a higher cost due to its pay-per-token pricing of $0.12 for inputs and $0.37 for outputs. Its benchmark score of 57.2 across four independent tests indicates strong performance, though the coverage is moderate. Gemma 4 31B is ideal for projects requiring extensive input handling and diverse output capabilities within a budget that allows for higher costs.