Node Workflow Studio Spaces

What Is the Node Workflow Studio Spaces?

An executable local text/data pipeline, distinct from a diagram maker. Connect text inputs, transformations and outputs on a draggable node canvas. Thirty node types include document text extraction, image/video/audio previews, generic file metadata, trimming, deduplication, sorting, case conversion, literal replacement, line filtering, prefix/suffix, merging, JSON formatting/path extraction, Unicode Base64, statistics, numbers, toggles, list selectors, templates, line slices, four-way routing, connected AI text models and custom routing nodes.

How It Works

Choose from ten starting templates or add nodes from the library. Use Groups to select members, label and color each frame; drag the group label or use arrow keys to move unlocked members together. Shift-click node headers to select members. Group settings survive JSON backups. Wider media cards preserve portrait and landscape previews with native video controls. Assets offers PDF, DOCX and text extraction; images, video and audio preview locally, with PNG frame capture for video. File Inspector accepts any file for metadata without executing it. Document outputs contain extracted editable text; media outputs contain metadata only, not media bytes. Reattach media after loading a backup. Drag headers to reposition nodes, then click or drag from an output port to an input port. Edit text on each card; use its gear for description, outline color, lock, bypass and processor settings. The side library pushes the canvas instead of covering it. Custom nodes support up to 128 named, colored inputs and outputs per side; use the gear to merge inputs or map each output to an input. Port names appear on hover or keyboard focus, and project JSON preserves the configuration. Run Workflow executes in dependency order; click a node for its full output. Fit and Arrange organize the canvas. Choose curved, straight or stepped wires, grid snapping, a minimap and fullscreen. Run a single branch from its node; inspect recent execution summaries in History. Save Locally stores one workflow, while JSON export/import transfers editable projects. SVG exports a diagram snapshot; PDF exports a text/results report.

When to Use It

Build reusable cleanup chains, compose a structured brief from multiple inputs, or extract data from JSON without repeating individual tools manually.

Frequently Asked Questions

Does it run AI models?

Yes, optionally. Connect your own OpenRouter API key and choose a text-output model in the AI node settings. Requests use your provider balance and send that node’s prompt and system instructions to the provider. Your key stays in memory and is not saved or exported. Local nodes need no account. Image/video generation and arbitrary code execution are not included.

How are connections validated?

Cycles and duplicate connections are rejected. Named ports vary by processor. Merge offers four slots (the first accepts multiple connections); replacement and prefix nodes accept connected settings; filters expose matched/unmatched branches and statistics expose separate counts. Per-node errors identify missing inputs or invalid data, and downstream failures are reported.

What are the limits?

Up to 100 nodes, 300 connections and 200,000 output characters per node. Documents: 10 MB and 200 PDF pages; images: 20 MB; other files: 100 MB, with a 200 MB session file budget. Scanned PDFs need OCR first; document layout is not preserved. Browser codecs determine media playback. Import supports BrowserTools JSON projects under 5 MB. Export sensitive inputs only to files you intend to keep private.

Found an issue? Send Feedback. See our data sources and update policy for tools that depend on changing information.