PearAI 🍐 Open-Source, Extendable AI Code Editor

The best AI coding tools - natively integrated into one super editor.

Nang Ang

3 months ago

#ai#culture#open_source#developer_tools

Hi, we’re Pan and Nang, and we’re making an open source AI code editor. The first time you may have seen us, you saw our MVP. Today, here’s our V1, and how we’re building the future of coding with AI.

TL;DR:

AI is transforming how coding is done. Developers want to leverage AI as much as possible, but can’t keep up with the evolving landscape of available tools. PearAI offers an open-source solution with up-to-date, curated inventory of the best AI tools, natively integrated for effortless AI-powered coding.

Demo

https://youtu.be/UpxhRnKj5Ck?si=P0xsVY8COvKF7Tqe

The Problem:

For developers:

“AI is transforming how development is done and I want to increase my productivity to the max using AI”

I don’t know which are the best AI tools to use
I don’t want 5+ different subscriptions and having to keep up with all of them and new ones
User experience with different tools is clunky and fragmented
I don’t know how to integrate them into my workflow and fully utilize them

For AI coding tools:

“I built the best AI tool for a use case. I want it to be integrated into a user’s workflow and to distribute to more users.”

Difficult to have tools communicate with other tools in the AI tech stack.
Churn is high because users don’t know how to integrate the tool into workflow
Distribution is hard and cost per acquisition is high

Current AI code editors:

Not open source, this leads to
- Lack of trust and transparency from big companies.On developer tooling teams that Pan and Nang have worked at, onboarding Cursor was not possible due to this. An open source developer tool would be much more favorable.
- Lack of community and flexibility; a code editor is the best product to be open source because developers actually use the product and know best how to improve it.
Try to solve coding with AI as a one-size-fits-all solution. For each specific part of coding with AI (ie. autocomplete, RAG, search, deploy, personalization, etc.), there exists another product that specializes in solving this single part - and does it better. This trend will only continue in the future. Check out the YC batch every season.

The Solution

An extendable, open source AI code editor. Imagine if VSCode was made knowing that AI was going to transform how development is done.

For developers:

Native integrations for the best user experience. A single unified UX for all AI tools.
Single subscription for access to all tools
Active open source community, helping with new features and customer support, led by build-in-public founders
PearAI curates an inventory of the best AI tools on the market.

For AI Tools:

PearAI integrates the tool into a unified, seamless user experience/workflow that users know and love. Here’s exactly what we mean:
- Functionality that users want from the tool is kept intact.
- The AI tool can communicate with other AI tools within PearAI if needed.
- Native integration into PearAI’s UX is added for AI tools.
- We uphold bring-your-own-API-key on tools that support it.
- AI tools gain more visibility and users. We currently have Continue, aider, mem0, and Perplexity as integrations under a single UX. As always, it’s made extremely clear that the user is using the underlying tool, and that they can (and should!) use the tool independently if they so choose.
- If an AI tool is paid, the company can charge using their API key or preferred plan.
- The AI tool will have the option to gain deeper insights and usage data from users within PearAI.

Our Vision:

Currently, we have hit a checkpoint in our product, where layers of the tech stack (i.e. Search, Chat, Code Generation etc.) can communicate with each other under a single, beautiful user experience our users love. This is our V1, and we’re just getting started. Building for the future, we envision PearAI as a flexible framework for AI tools to live and integrate. Not only will the number of layers of the AI tech stack grow, but the number of companies solving each layer of this AI tech stack will grow as well. As a result, the quality of each product for every part of coding with AI will improve. The best user experience can be achieved if these AI tools have a framework to communicate with each other, and users have access to them in a single, unified IDE. With PearAI, we’re hoping to build a win-win ecosystem where users can make what they want, while AI companies gain expanded reach and deeper insights.

ABOUT US:

🟢 Nang Ang, B.S.&M.S @ CMU in Computer Engineering, previously software engineer at IMC Trading, Coinbase, and Two Sigma on the Developer Tooling team. He also has been sharing his coding journey and technical projects, and has grown a community of over 150k subscribers on YouTube, with over 30 million total views on social media from developers.

🔵 Duke Pan, prev software engineer at Meta, Cisco, Coinbase, and Tesla Autopilot. He also has been creating stories and making content since he was 8 years old, growing a community of over 345k subscribers on YouTube, and gaining over 50 million total views from mainly people in tech and developers.

We are building fully in public on social media with our wonderful open source community.

Our Ask:

Check out our product at: https://trypear.ai - we’re sure you’ll love it. :) May the best AI code editor win!