Gemini 3 enables multimodal function calling via Interactions API, allowing tools to return images processed natively.
Read the original at www.philschmid.de→Multimodal function calling allows tools to return images the model can process natively, similar to how you pass images in prompts. Instead of describing what's in a file, your tool returns the actual image and...
Original headline: "Multimodal Function Calling with Gemini 3 and Interactions API"