tscircuit is a TypeScript/TSX framework that enables AI agents to design and generate physical circuit boards. It provides a React-like component model for electronics, real-time autorouting, and native AI coding skills, allowing any AI to create manufacturable PCBs. The platform is open-source under MIT and includes features like visual diffs in GitHub and KiCad import.
Funding
Funding not disclosed
Founders
Product
Problem
Traditional electronic design automation (EDA) tools are complex, desktop-bound, and require deep expertise, creating a steep learning curve and slowing down the design process. This complexity also makes it difficult for AI agents to generate or modify circuit boards, limiting the potential for automated hardware design and rapid prototyping.
Solution
tscircuit is an open-source TSX framework that reimagines circuit design as a component-based, code-driven process. It allows developers and AI agents to compose circuit boards using familiar React-like syntax, with elements such as resistors, chips, and capacitors as components. The platform includes a sub-second autorouter for 4-layer boards, runs locally, and provides an AI "skill" that gives agents the CLI, syntax, and workflow context needed to make useful hardware changes. tscircuit also integrates with version control, enabling visual diffs, approvals, and rollbacks directly within GitHub, and supports importing from popular libraries like KiCad and JLCPCB.</problem> <solution>The core value proposition is making electronics generation accessible and automatable. By treating circuit design as code, tscircuit enables any AI agent to create, validate, and iterate on physical boards without special installation. The platform provides a full suite of built-in elements, from basic components to complex features like differential pairs and analog simulation, and generates standard fabrication files for manufacturing. This approach reduces the barrier to entry for hardware design and accelerates the development cycle from concept to physical prototype.
Target Audience
Primary users are hardware engineers, software developers, and AI researchers who want to leverage AI for automated circuit design, as well as teams building custom electronics who need a fast, code-based workflow.
Features
- TSX component model for circuits, allowing composition with props, children, and hooks for electronic elements.
- Real-time autorouting with sub-second performance on 4-layer boards, designed for LLM-compatible tweaking.
- Native AI coding skills for agents like Claude Code and Codex, providing CLI, syntax, and pre-fabrication context.
- Visual diffing and version control integration inside GitHub for approvals, comments, and rollback.
- Open-source (MIT) with a comprehensive library of built-in elements, including <chip />, <resistor />, <capacitor />, and <analogsimulation />.
- Import capabilities from KiCad, JLCPCB, and Ultra Librarian, plus support for Circuit JSON and custom TypeScript libraries.