NodeFlow.jsx
Visual node editor
Build complex multi-modal pipelines on a drag-and-drop canvas. Color-coded paths route text, image, video, and audio between nodes.
NodesNavbar.jsxDark / light themes
Visual workflow builder
Orchestrate multi-modal AI generation
Design drag-and-drop pipelines for text, image, video, and audio — then run them in the studio or trigger headless workflows via REST.
100+
AI models
4
Modalities
REST
API ready
Open source
npm run workflow:devWorkflow canvas
Color-coded routing across every modality
TextImageVideoAudio
Prompt Input
Text
UnifiedNode
Image
Video Render
Video
Video Combiner
Video
Audio Node
Audio
Output
Payload
TIVA
NodeFlow.jsx · React Flow
Live previewEverything you need
Built for visual orchestration
From canvas editing to conversational planning — one studio unifies the full generative stack for creators and developers.
ChatWidget.jsx
AI Architect Assistant
Conversational widget powered by a FastAPI LLM planner. Generates, builds, and revises node graphs from natural language with Markdown logs.
workflow_router.py
> planning workflow…
Generated 5 nodes · text → image → video
Template saved
UnifiedNode.jsx
Unified node system
Text, Image, Video, and Audio modes with RenderField schema inputs, NodeSendButton execution, NodeOptionsMenu, and UploadNode.
RenderField.jsxUploadNode.jsx
Utility nodes
Composition & preview
PromptConcate merges text streams. VideoCombiner links clips. AudioPlayer and VideoPlayer preview generated content inline.
PromptConcate.jsxVideoCombiner.jsx
local_workflow_engine.py
FastAPI orchestration
Async node execution with list-variable mapping, status polling, and storage. REST APIs to save, template, publish, and trigger workflows.
Headless REST
useGenerationCost.jsx
Live cost estimation
Monitor runtime resource consumption during pipeline execution with sharp metrics and cluster utilization tracking.
Production monitoring
Use cases
For creators, developers, and teams
Generative AI creators
Chain text prompts into images, upscale frames, render video, and attach audio — or produce complete short films with music, voice, and lip-sync.
Developers & integrators
Publish workflows as templates and trigger them via REST endpoints. Design custom UI nodes with schema-driven validation bindings.
Content producers
Automate recurring creation with reference templates. Monitor live generation cost as pipelines execute.
Headless REST automation
FastAPI orchestration engine
Publish canvas assets as programmatic templates. Trigger pipelines, poll statuses, and configure list-variable inputs via workflow_router.py.
Production cost allocation
Track cluster utilization with live runtime cost from useGenerationCost.jsx.
workflow_router.py
@router.post("/workflows/{workflow_id}/run")
async def run_workflow(
workflow_id: str,
payload: WorkflowRunRequest,
engine: WorkflowEngine = Depends(get_engine),
):
job = await engine.enqueue(
workflow_id=workflow_id,
inputs=payload.inputs,
poll_interval=payload.poll_interval,
)
return WorkflowStatusResponse(
job_id=job.id,
status="running",
storage_uri=job.output_uri,
)useGenerationCost.jsx
Live runtime cost
Tokens
128.4k
GPU min
14.2
Jobs
3 active
$0.08
Text
$1.24
Image
$4.80
Video
$0.32
Audio
Polling cluster utilization every 2s