run_api_node
Run any ComfyUI node in one call: auto-builds the workflow, inserts needed loaders/save nodes, and works with all 200+ API nodes.
Instructions
Run any ComfyUI node in a single call.
Builds a minimal workflow around the node, auto-inserting loader and save nodes as needed. Works with all 200+ API nodes (Kling, Runway, Luma, Stability, ElevenLabs, Gemini, etc.) and any built-in node.
Workflow: discover nodes with list_api_nodes() or search_nodes(), check inputs with get_node_schema(), then run with this tool.
For IMAGE/VIDEO/AUDIO-typed inputs, pass a filename string (e.g. "photo.png") and a LoadImage/LoadVideo/LoadAudio node is auto-inserted.
Input Schema
| Name | Required | Description | Default |
|---|---|---|---|
| inputs | No | JSON string of node inputs. Default: "{}". Primitive values: {"prompt": "a cat", "seed": 42}. File inputs: {"image": "photo.png"} — auto-creates LoadImage. Use get_node_schema(node_class) to see all available inputs with their types, defaults, and valid ranges. | {} |
| timeout | No | Max seconds to wait for completion. Default: 0 (use server default, typically 300s). Set higher for slow API nodes like video generation (e.g. 600 for Veo 3). | |
| node_class | Yes | The node class name (e.g. "KlingTextToVideoNode", "OpenAIDalle3", "ElevenLabsTextToSpeech"). Use list_api_nodes() or search_nodes() to discover available nodes, and get_node_schema() to see required/optional inputs. | |
| queue_only | No | If true, queue the job and return immediately with the prompt_id without waiting for completion. Default: false. Use get_job_status(prompt_id) to check progress later, or wait_for_jobs([prompt_id1, prompt_id2, ...]) to wait for multiple jobs at once. | |
| output_format | No | How to save the output. Default: "auto". "auto" — detect from node outputs (IMAGE→SaveImage, etc.). "image", "video", "audio" — force a specific save type. "none" — skip save node (for nodes that save internally). | auto |
Output Schema
| Name | Required | Description | Default |
|---|---|---|---|
No arguments | |||