Image input / vision

Image input (also called vision) is Claude's ability to receive, read, and reason about static image files alongside text. When you attach a photo, screenshot, chart, diagram, or document image to a conversation, Claude analyzes the visual content and responds in natural language — describing what it sees, extracting text, interpreting data, comparing designs, or answering questions about the image. This is a one-way capability: Claude can understand images but cannot generate, edit, manipulate, or create them. Vision works across all current Claude models (Claude 3 and Claude 4 families) and is available on every plan tier — Free through Enterprise. You can include multiple images in a single turn; on claude.ai the limit is 20 images per turn (10 MB each), while the API supports up to 600 images per request subject to overall payload size limits. Practical applications range from simple tasks like reading a receipt or describing a photo, to intermediate work like interpreting financial charts or reviewing code screenshots, to advanced workflows like analyzing multi-page PDFs containing embedded charts, converting wireframe sketches into code, or powering document-processing pipelines in enterprise knowledge bases.