04 · Multimodal input

Vision

GPT-6 models can read images as well as text. Upload a photo, screenshot, handwritten note or chart and ask about it. The image is sent as base64 in the same message as your question. Docs ↗