Build typed AI pipelines visually or from an agent, then run them anywhere, from the API, SDK, MCP, or CLI.
curl -N -X POST https://studio.blitflow.com/api/v1/runs \
-H "Authorization: Bearer $BLITFLOW_TOKEN" \
-H "Content-Type: application/json" \
-d '{"workflowRef":"acme/sprite-pack@2.1.0"}'Drag nodes onto an infinite canvas and wire them together. Every change saves as you work.
Publish once and call it from your app. Every feature gets its own versioned workflow behind one stable contract.
Publishing freezes a workflow and pins every node it uses to an exact release.
From the first node to production traffic, in one place.
Describe a pipeline and an agent assembles the graph.
Ports only connect when the types line up.
Publish one and it becomes a node in the next graph.
Every step reports its cost as it completes.
One dashboard for every run and what it cost.
Invite your team and gate who can change what.
One typed contract, so every surface stays in lockstep.
Call a workflow over REST and stream the run back.
Await the outputs, or follow events as they land.
Hand your workflows to an agent, hosted or local.
Run a workflow and watch it from your terminal.
Every category behind one typed node interface, so a video model wires up the same way as a text model or a resize utility.