Documentation

LoopLlama API

LoopLlama is the orchestration API for teams of AI agents. Define a workflow, give it a crew of agents, and trigger runs over a simple REST API — we handle the sequencing, model calls, persistence, and usage metering.

How it works#

A workflow bundles a reusable configuration: a name, a default model, and a crew of one or more agents. Each agent has a role and a system prompt. When you trigger a run, LoopLlama executes the crew sequentially — each agent sees the original input plus the prior agents' output — and records every step with its tokens and timing.

  • Workflows — reusable agent configurations you create once and run many times.
  • Crews & agents — the ordered list of roles that process each run, from a planner to a writer to a reviewer.
  • Runs & steps — a single execution of a workflow, broken down into one step per agent.

Base URL#

All API requests are made to the versioned base URL over HTTPS. The current version is v1.

Base URL
https://api.loopllama.ai/v1

Running a self-hosted or local instance? The same routes are served under /api/v1 on your deployment — for example http://localhost:3000/api/v1.

Make your first request#

Every request is authenticated with a Bearer API key. Create one in the dashboard, then list your workflows:

bash
curl https://api.loopllama.ai/v1/workflows \
  -H "Authorization: Bearer $LOOPLLAMA_API_KEY"

Explore the docs#