Prebuilt workflow · Generate assets

SVG Generator

Generate clean, production-ready SVG markup from a prompt and an optional reference image.

· You bring Prompt + reference· output.svg

IMAGE
Reference Image (Optional)
Reference Image (Optional)4:3
IMAGE
SVG Generation
GPT 6 AstraHigh

Turn the sketch into Fernway's four trail-difficulty badges (easy, moderate, hard, expert), laid out in one row: rounded hexagons with a single glyph each, forest #1F3A2E strokes on chalk #F4F1EA, the expert flag in signal orange #F26B3A. Create clean, production-ready SVG artwork from this prompt and optional reference image. Requirements: - Return valid standalone <svg>...</svg> markup - Use vector primitives and paths only (no raster output) - Keep the structure readable…

10 lines · 604 charsEditorExpand
AttachmentsResponse
SVG$0.12Run
COMPOSEsvg
output.svg
output.svg33 lines · 1.4k chars
CODE
Run all

A real run, unedited1 generation · $0.12 of walletBrand and inputs are fictional, made in Vmotif · 2026-09-24

What came back

Straight off the canvas.

Everything below is what this run returned, untouched.

How it runs

One step, one prompt.

What each stage reads, what it makes and what it cost on this run.

  1. 01

    SVG Generation

    Reads your inputsMakes 1 SVG

    1 stepGPT 6 Astra$0.12

From your agent

Or let your agent run it.

This run went through the Vmotif MCP server, the way Claude Code, Cursor or any MCP client drives it: seed the canvas, fill the inputs, run the graph.

create_canvas({ premade_workflow_id: "svg-generator" })
update_canvas({ canvas_id, updates: [/* your inputs */] })
run_workflow({ workflow_id: canvas_id })
MCP docs
Where it fits

Part of a bigger job.

More in Generate assets

Produce finished visual assets — vectors, video, and bulk style variations — straight from the canvas.