01Core
Reliable multi-agent orchestration
Coordinate tasks across agents with automatic retries, failure handling, and execution ordering. You define what each agent does — awg handles how they wire together.
DEVELOPER INFRASTRUCTURE · AI AGENT ORCHESTRATION
Reliable multi-agent infrastructure for individual developers. awg handles task handoffs, retries, and execution flow — so you can focus on what your agents actually do.
HOW IT WORKS
Define your team, declare dependencies, and POST to /v1/runs. awg resolves execution order, monitors agent health, and retries failures automatically. You get a single run_id back immediately.
POST /v1/runs
{
"team_id": "team_42",
"agents": [
{ "id": "agent.planner", "depends_on": [] },
{ "id": "agent.research", "depends_on": ["agent.planner"] },
{ "id": "agent.writer", "depends_on": ["agent.research"] }
]
}
// → 200 OK run_id: run_8f3aCAPABILITIES
Coordinate tasks across agents with automatic retries, failure handling, and execution ordering. You define what each agent does — awg handles how they wire together.
Plug in from LangChain, CrewAI, AutoGen, or plain API calls. awg sits at the orchestration layer, not the model layer — it works with whatever you're already using.
Free tier for individual experimentation. Paid tiers unlock higher concurrency, run history, and reliability features. You pay as your agent systems grow — not upfront.
EARLY ACCESS
Free tier available. Usage-based pricing as your agents scale. No upfront commitment.
Start building for free →