The official app is free to use, and its source is available for study and upstream contributions under the AIBoard Community Source License. Independent redistribution or hosting requires permission. You pay your AI providers directly for the tokens used.

Which API keys do I need?

At least one provider key or connected account. Chat compares two or more models; a build needs an architect and at least one worker, which can be the same model. Local models via Ollama or LM Studio need no key at all.

Can I use it offline?

The app loads as a static site, but chats and builds need network access to reach the AI providers — unless you use only local models on your own machine.

Can I use my ChatGPT subscription instead of an API key?

Yes. Connect a ChatGPT Plus/Pro or GitHub Copilot account through the local account runner and its models appear alongside your API-key models — in chat and in Build mode. Rate limits and availability are governed by your provider account.

About AI Board

A local-first AI harness: chat and build with the models you choose, from the providers you already use.

Chat

Ask one question and several models work it out together. They answer, read each other's responses, and refine or challenge their positions over a few rounds. A judge model you pick then writes one answer with a confidence score and any notable dissent.

Panel — models build on each other toward a shared answer.
Debate — models argue opposing positions so the trade-offs come out.
Specialist + reviewers — one model drafts, the others review and demand fixes.

Build

Describe a change and a crew carries it out on your own repository. An architect plans the work as tasks, workers implement them in isolated Git worktrees, and an independent verifier checks the evidence before anything is accepted. You approve the gates that matter and choose how the finished work is handed back.

Builds run on Runner V2, a native agent kernel you start on your machine. It owns files, Git, shell commands, checkpoints, and recovery — the browser never executes project code.

Any provider

OpenAI, Anthropic (direct or Azure Foundry), Google Gemini, OpenRouter, xAI, Meta, NVIDIA NIM, and any OpenAI-compatible endpoint such as Ollama or LM Studio. ChatGPT Plus/Pro and GitHub Copilot accounts work too, through a small local account runner. Mix them freely: a strong architect with inexpensive workers is a common build crew.

The Benchmark measures which models and teams work best for you — tool use, team lift, and real build tasks — using your own keys.

Local-first and private by design

There is no backend. The app is a static site that calls the AI providers directly with the keys you enter in Settings. Keys, sessions, and attachments stay in your browser (IndexedDB) or in a local folder you pick, optionally encrypted with a passphrase using AES-256-GCM. Nothing is uploaded to a server of ours, because there isn't one.

Use an AI account you already pay for

If you subscribe to ChatGPT Plus/Pro or GitHub Copilot, you can point AI Board at that account instead of entering per-token API keys. A small account runner runs on your machine, you authorize it once with OpenAI or GitHub, and the app talks to it over localhost — the login stays in a file on your computer.

A connected account works everywhere the app uses models: chat, builds, and the benchmark.

This draws on your account's own entitlement, so rate limits, model availability, and behavior are set by OpenAI or GitHub rather than by AI Board.

Author and license

The current project is source-available under the AIBoard Community Source License. You can inspect the code and submit improvements upstream, but independent distribution and hosted forks need written permission. Earlier AGPL-licensed copies retain their original rights.

Contributions follow the contributor agreement. AIBoard branding is governed by the trademark policy.

Contact

For feedback, bug reports, privacy questions, or collaboration, email [email protected].

Source code is available on GitHub.

FAQ

Is it free?: The official app is free to use, and its source is available for study and upstream contributions under the AIBoard Community Source License. Independent redistribution or hosting requires permission. You pay your AI providers directly for the tokens used.
Which API keys do I need?: At least one provider key or connected account. Chat compares two or more models; a build needs an architect and at least one worker, which can be the same model. Local models via Ollama or LM Studio need no key at all.
Can I use it offline?: The app loads as a static site, but chats and builds need network access to reach the AI providers — unless you use only local models on your own machine.
Can I use my ChatGPT subscription instead of an API key?: Yes. Connect a ChatGPT Plus/Pro or GitHub Copilot account through the local account runner and its models appear alongside your API-key models — in chat and in Build mode. Rate limits and availability are governed by your provider account.

Open the workspace