Stable Diffusion image info reader
Read the prompt, negative prompt, seed, sampler and model from a generated PNG — WebUI, ComfyUI and NovelAI — and export the ComfyUI workflow.
Or choose them. Files are processed in this tab and never uploaded to any server.
Images made with Stable Diffusion WebUI, Forge, ComfyUI or NovelAI carry their own recipe: the prompt, negative prompt, steps, sampler, CFG scale, seed, model hash and LoRAs are written into the PNG's text chunks, and ComfyUI even stores the whole node graph. This reader does what WebUI's PNG Info tab does, in your browser: drop the image and get the parameters parsed into fields, the raw chunks for anything unusual, and — for ComfyUI images — the workflow as a .json you can drag back onto the canvas. JPEG and WebP exports from WebUI are read from their EXIF comment too.
How it works
- tEXt, zTXt and iTXt chunks are read directly from the PNG without decoding the image, so a 50 MB upscale is parsed instantly.
- WebUI's parameters string is split into prompt, negative prompt and the settings line, with quoted values and the JSON hash blocks kept whole.
- ComfyUI's prompt graph is summarised to the human-relevant nodes — text encoders, sampler, checkpoint, latent size, LoRAs — and the workflow chunk becomes a downloadable file.
Where your data goes
Nowhere. This tool runs entirely in your browser: the text you paste is processed by the page and is never transmitted to a server or written to a log.
This tool is free and needs no account. Its results exist only in your open page and are not saved anywhere.
What it costs
This tool is free, with no sign-in and no points.
Common questions
- The image shows no metadata. Why?
- Most platforms strip it: Discord, X, Reddit, Instagram and WeChat re-encode uploads and drop text chunks, and screenshots never had any. Only the original file saved by the generator keeps the parameters. Some generators also have a setting to not write them.
- Can it read images from Midjourney, DALL·E or Flux?
- Midjourney and DALL·E do not embed prompts in a readable form. Flux images made in ComfyUI or Forge carry the same chunks as any other model and read normally; what matters is the front-end, not the model. Fooocus and SD.Next use the WebUI format and work too.
- What exactly is in the ComfyUI workflow file?
- The full graph as saved by ComfyUI — nodes, links, positions and widget values — which is what you drop onto the canvas to restore it. The separate prompt chunk is the executable API form of the same graph, which this tool summarises for reading rather than exporting.
The open-source behind it
This tool is a self-contained implementation. AUTOMATIC1111/stable-diffusion-webui (AGPL-3.0) does the same job as a library — if you need this behaviour inside your own program, start there rather than calling a web page.
AUTOMATIC1111/stable-diffusion-webuiAlso known as
- png info stable diffusion
- read prompt from image
- comfyui workflow from png
- sd metadata viewer
- automatic1111 parameters png
- image generation metadata reader