Gen nodes
Send context into model calls.
Feed text, image, and structured context into generation nodes without rebuilding the prompt by hand each time.
Visual AI workflow editor
Build AI workflows on a graph canvas.
ThreaderGen is for multi-step generation work: wire inputs into text, image, script, voice, and structured-output nodes, then save the workflow as a reusable project that can run from the command line for fully automated systems.
Typed
connect text, image, audio, video, and structured data
Nested
wrap reusable graphs with explicit inputs and outputs
Runnable
run projects from the editor or project JSON
Workspace
A node editor for generation workflows.
Use the canvas to connect source material, prompts, model calls, generated assets, structured records, and custom scripts. The graph shows where each output came from and what depends on it.
Structured data
Inspect structured outputs.
Structured nodes make JSON-like records visible, editable, and reusable by downstream nodes.
Subprojects
Reuse subgraphs.
Mark nodes as project inputs or outputs, then call that graph from another graph as a subproject.
Iteration
Run a workflow across many inputs.
Iterate mode runs a child graph across array inputs and returns the collected outputs for review or the next node.
Managed surfaces
Supporting tools are built in.
Schema library
Define response shapes once and reuse them across Gen Text nodes.
Voice catalog
Store named ElevenLabs voices for audio generation workflows.
Command-line runs
Run saved graphs from the command line so editor-built workflows can move into automation.
Script and agent nodes
Add code and agent work as explicit nodes instead of hiding them outside the flow.
