Input Components
Input components (SmartHopper → B. Input) package Grasshopper data, files, web pages, forum posts, or system instructions into an Input > (>) wire that any Output component can consume.
Primitives & lists
Single-value inputs package one item per value; * List to AI components collapse an entire list (branch) into a single AI input so the model sees the whole list in one call.
| Icon | Component | Nickname | Packages | Preview |
|---|---|---|---|---|
| Text to AI | Text2AI | A single text prompt or message | ![]() | |
| Text List to AI | TextList2AI | A list of strings as one JSON array per branch | ![]() | |
| Number to AI / Number List to AI | Num2AI / NumList2AI | A single number, or a list of numbers per branch | ![]() | |
| Integer to AI / Integer List to AI | Int2AI / IntList2AI | A single integer, or a list of integers per branch | ![]() | |
| Boolean to AI / Boolean List to AI | Bool2AI / BoolList2AI | A single boolean, or a list of booleans per branch | ![]() | |
| JSON to AI | Json2AI | Validated JSON text | ![]() |
Media, files, web & canvas
| Icon | Component | Nickname | What it packages | Preview |
|---|---|---|---|---|
| Image to AI | Img2AI | An image (Bitmap, file path, URL, data URI, or base64) for vision-capable models. | ![]() | |
| Audio to AI | Audio2AI | A local audio file (.mp3, .wav, .m4a, .ogg, .flac) for audio-capable models. | ![]() | |
| File to AI | File2AI | Local documents (PDF, DOCX, XLSX, PPTX, HTML, CSV, JSON, XML, TXT, EML, EPUB, RTF…) converted to Markdown, with embedded images materialized into the payload for vision-capable models. Also outputs the converted Markdown and detected Format. | ![]() | |
| Web to AI | Web2AI | Web page URLs fetched and converted to Markdown, with page images materialized into the payload for vision-capable models. Also outputs the extracted Markdown. | ![]() | |
| Canvas to AI | GhJSON2AI | The active Grasshopper canvas (or selected/filtered components) exported as GhJSON. Also outputs the GhJSON string. | ![]() |
Forum posts & topics
Fetch a single post (by Post ID) or a whole thread (by Topic ID, with optional Max Posts) and output both Input > and the raw Post JSON / Topic JSON:
| Icon | Component | Nickname | Source | Preview |
|---|---|---|---|---|
| McNeel Post to AI / McNeel Topic to AI | McNeelPost2AI / McNeelTopic2AI | McNeel Discourse forum | ![]() | |
| Ladybug Post to AI / Ladybug Topic to AI | LadybugPost2AI / LadybugTopic2AI | Ladybug Tools forum | ![]() | |
| Discourse Post to AI / Discourse Topic to AI | DiscoursePost2AI / DiscourseTopic2AI | Any Discourse forum (pass Base URL) | ![]() |
Prompt and context inputs
| Icon | Component | Nickname | What it does | Preview |
|---|---|---|---|---|
| AI Prompt | AIPrompt | Packages a system prompt (instructions that set the AI's role, tone, or rules) into Input >. | ![]() | |
| AIContext | AICtx | Retrieves live data from a registered context provider by Provider ID (e.g., time, file, environment, selection, viewport) and outputs both Input > and a human-readable Display string. Use Context Providers (A. AI) to list available IDs. | ![]() |
Combining multiple inputs
Wire several input components into the same Input > parameter on an Output component (hold Shift while wiring, or use Grasshopper's Merge) to combine them into a single AI call:

Full files, web pages, and unfiltered canvas exports (Include Runtime Data = True) can be large. Filter canvas exports or summarize long documents first when working with smaller context windows.















