curl https://api.orchestracode.dev/v1/tasks \
-H 'authorization: Bearer $ORCH_KEY' \
-H 'x-github-token: $GH_TOKEN' \
-d '{
"repo": "you/your-app",
"branch": "main",
"task": "Add a dark-mode toggle to the settings page",
"model": "claude-sonnet-4-6"
}'
# → { "task_id": "...", "pr_url": "https://github.com/you/your-app/pull/42" }Cursor and Claude Code do beautiful work when you're sitting there. Close the laptop and the run dies. A background AI coding agent hosts the loop, the sandbox, and the git push in the cloud, so the agent finishes whether you're online or asleep.
Stateless POST. The background coding agent streams every tool call, edit, and commit over SSE, so your dashboard renders progress live. The branch lands on origin under your token. We never store it.
Pass your Anthropic, OpenAI, or open-weights key on each request. We never store it. Model tokens bill directly to your provider, never marked up.
A background AI coding agent should cost like an API call, not like a $500/month seat. Pay for the agent loop, the sandbox, and the git push. That is it.
Most background agents are GitHub-only. We push pull requests to both, with the same payload shape and the same per-request token model.
Every plan is BYOK. Your Anthropic, OpenAI, or open-weights key. We charge for the agent loop, the sandbox, and the git push. Model tokens are billed by your provider, never marked up.
For trying it out on a side project.
For one developer shipping serious work.
For small teams building with the API in production.
For high-volume API embeds and enterprise contracts.
A background coding agent only pays off if your team actually has work to hand off. Tell us what kind.