100% In-Browser · Zero-Upload Edge Privacy · RFC 2083 PNG Standard

ComfyUI PNG Metadata Injector

Inject execution graphs and prompt metadata directly into any image file. Download native ComfyUI PNGs that drag & drop directly into your canvas workspace.

Step 1: Select Target Image PNG / JPG / WebP

Click or drag & drop image here

Supports PNG, JPG, or WebP. Converted to native lossless PNG in-memory.

Or Pick Curated Sample Artwork:
Target image preview
Resolution: Auto Format: PNG

Step 2: Choose ComfyUI Workflow JSON 28 Blueprints

OR PASTE CUSTOM WORKFLOW JSON
Selected Blueprint: Civitai #144303076 Snow White Window: Arched Sunlight & Porcelain Film Aesthetics
Architecture: Krea 2 Turbo (Qwen3-VL 4B + Wan 2.1 16-ch VAE)
Embedded Chunks: tEXt:workflow + tEXt:prompt (RFC 2083)

Step 3: Generate Native Drop-Ready PNG

Assembles PNG binary chunks, calculates 32-bit CRC checksums, and encodes metadata in client memory.

How ComfyUI Reads PNG Metadata

PNG files consist of an 8-byte signature followed by structured chunks (IHDR, IDAT, tEXt, IEND). ComfyUI inspects ancillary tEXt chunks for key strings:

  • tEXt:workflow : Serialized canvas graph with node coordinates and splines.
  • tEXt:prompt : Serialized execution DAG API format consumed by ComfyUI backend.
Action completed