PatentFig AI

About PatentFig AI

PatentFig AI generates patent drawings from a text description, a sketch, a product photo or a CAD screenshot, and targets the filing conventions of specific patent offices rather than producing generic technical illustration. It covers line art with reference numerals and leader lines placed to patent-office convention, 3D renderings with multiple view angles, and flowchart and block diagrams, with output aware of USPTO, EPO, CNIPA, JPO, KIPO and PCT requirements. The workflow is anchor-plus-multi-view: generate a base figure, then request the other views from it so the set stays consistent, which is the part that makes it a patent tool rather than an image generator. Figures are edited conversationally — 'make the arrow thicker', 'add a cross-section view' — and every generation creates a numbered version with sub-versions for edits, so a figure set can be compared and reverted rather than regenerated from scratch. Export covers PDF, PNG, SVG and TIFF at exact office specifications, with batch export of all views of a figure. Beyond the generator the site publishes a set of adjacent free tools — a figure compliance checker, format conversion, vectorisation and DPI enhancement — plus drawing requirements and standards references, templates and checklists, and a glossary. There is a documented developer API. The company states it does not train on customer data and encrypts in transit and at rest.

Key Features

Tags