Full workflow graphs
Reads the embedded node workflow and resolves the effective prompt, model, sampler and seed from the graph.
Drop a generated PNG below to see its prompt, negative prompt, model, sampler, seed and full workflow — parsed instantly in your browser. Your image is never uploaded anywhere.
Got a whole folder of these? Image MetaHub searches your entire local library the same way — for free.
Drag & drop a PNG here
or click to choose a file — nothing leaves your device
Image MetaHub searches your entire local library by prompt, model, LoRA and seed — instantly, without uploading anything.
The same parser that powers Image MetaHub's desktop library — reads embedded PNG metadata, no guessing.
Reads the embedded node workflow and resolves the effective prompt, model, sampler and seed from the graph.
Prompt, negative prompt, steps, CFG scale, sampler, seed, size, model hash and LoRAs.
Positive/negative prompt, model, VAE, scheduler and generation mode.
And Image MetaHub's own MetaHub Save Node payload for ComfyUI.
No. The PNG is read and parsed entirely in your browser using JavaScript. It's never sent to a server.
Stable Diffusion, ComfyUI, A1111, InvokeAI and Forge embed the prompt, model, seed, sampler and (for ComfyUI) the full node workflow directly inside the PNG's text chunks.
Some tools strip metadata on export, and social/messaging apps often re-encode images and remove it. Screenshots never had it to begin with.
ComfyUI can reopen the workflow from any PNG it saved — that's built in, not specific to any node. The MetaHub Save Node additionally writes clean, pre-resolved fields (prompt, model, seed, tags, notes, analytics) so tools don't have to reverse-engineer the graph.
Image MetaHub searches your whole local library by prompt, model, LoRA and seed — instantly, without uploading anything. Free to start.