Gemini 2.5 Computer Use Preview

JSON →
google multimodal
textimage

A preview model from Google's Gemini 2.5 series designed for computer use and automation tasks.

context window 128K tokens
max output 64K tokens
input price $1.25 / 1M tokens
output price $10 / 1M tokens
visiontool-usestreamingreasoningfunction-calling
releasedOct 2025