BellowAI documentation
BellowAI turns one idea into a complete delivery plan using a multi-agent LLM pipeline. These pages explain how to get started, how the system works, how to connect your own model keys, and how your data and keys are protected.
Start here
Getting started
Sign up, create a session, pick a model or add your own key, then run and review the pipeline output.
How it works
The path a run takes — from your browser through the API, a queue, an ephemeral worker, the LLM, and back as streamed results. Plus Parallel Agent Dispatch.
Bring your own keys
Connect Anthropic and OpenAI keys. The model picker shows only the models you have a key for.
Security
Authentication, two-factor, per-organization key isolation, and ephemeral compute.
What BellowAI does
You describe an idea in a prompt. A sequence of specialized agents takes it through planning, epics, features, stories, a development approach, and validation — each stage handing structured output to the next. You review and approve at every gate before work moves forward. For larger plans, independent slices of work can be dispatched to run in parallel and then merged with a conflict check.
You bring your own LLM key (Anthropic or OpenAI), so model usage is billed to you by the provider at cost, and you choose which models to use.