boogu / edit / editImage

Editing models are where the 2026 bases separated themselves, and Boogu ships one too. Civitai Image Gen (boogu / edit / editImage) is the instruction-edit operation: feed it an image, describe the change in plain language, get the edited result back as a native IMAGE. Same model, different job - this is the "change the background / make them smile / add an object" workflow.

What it is

The edit node in the Boogu family. Required: prompt - the instruction, which is the entire interface. Optional: images (an IMAGE socket for the source), plus the usual output_format, image_metadata, negative_prompt, sampler/scheduler, steps (default 25 - a touch higher than the create node), cfg_scale (default 3.5), seed, quantity (1–4), loras, diffusion_model (CIVITAI_AIR), width/height, api_config.

Note the quantity cap: 4, not 12. The edit engine batches tighter than Boogu's create path, and each output is billed.

Outputs: images, errors, workflow_id, raw_json.

How it works

The source image rides along with the job to Civitai's fleet; the model re-renders it against your instruction, and the pack converts the returned blob back into your graph. There's no denoise_strength dial here - like the other modern edit nodes, the prompt is the strength control. Be specific about what changes and what must stay: "make the background a rainy street" plus "keep the subject exactly as-is" reads better than a one-word wish.

Install & gotchas

Standard pack install: Manager → Civitai Comfy Nodes (publisher civitai), or comfy node registry-install civitai-comfy-nodes, or git clone + pip install -r requirements.txt. Auth via env var / OAuth / Civitai Auth node.

The things to keep straight:

1. Your source image leaves the machine. The edit input is uploaded with the job - the standard cloud-node data flow, worth remembering for anything sensitive or licensed.
2. Edit outputs are billed per image. The 1–4 quantity range exists partly to protect your Buzz; check raw_json if you want the exact cost per run.
3. It's a young model with a thin paper trail. Boogu's edit quality hasn't been stress-tested by the community the way Flux 2's has. Expect good results on straightforward changes and less certainty on surgical ones.

Honest take: for one-off edits on a GPU-less machine this node is genuinely useful, and Boogu's speed makes it a comfortable cloud citizen. But if edit quality is the whole game for you, the pack's Flux 2 Dev edit node has a far longer community track record - reach for that one when the edit has to be right.