Parchi

Parchi is an AI-powered browser sidebar agent designed to automate web interactions through natural language commands. It enables users to read web pages, fill out forms, click buttons, and navigate browser tabs without manual intervention. The tool is positioned as a browser copilot, streamlining routine web tasks and making browser automation accessible directly from the sidebar.

Among its capabilities, Parchi supports multi-model AI integration, allowing users to switch between providers such as Anthropic, OpenAI, OpenRouter, Kimi, or any compatible endpoint even in the middle of a conversation. It offers full browser control, including reading pages, interacting with elements, filling forms, navigating tabs, and taking screenshots. Vision-native features enable Parchi to analyze automatic screenshots using vision models, interpreting web pages as they appear visually to the user. Users can also save model configurations as profiles and record browser flows as replayable skills, enhancing automation and efficiency for repeated tasks.

Parchi is delivered as a browser extension, with installation available for Chrome. It is fully open source under the MIT license, allowing users to self-host the backend, audit the source code, and extend its functionality with custom tools. The platform's pricing model is flexible: users can bring their own API keys for free, enabling unlimited usage of all browser tools and AI providers, or opt to purchase prepaid credits for instant, keyless access on a pay-as-you-use basis, with no subscriptions required.

This tool is suitable for individuals or professionals seeking to automate browser-based workflows, leverage multiple AI models, or customize and extend browser automation within an open-source framework.