Stability Core Image Generation

If you're going to try any node in this pack first, make it this one. StabilityImageCore is the plain text-to-image workhorse: prompt in, image out, no local checkpoint to download, no VRAM to budget, no sampler to argue with. It calls Stability's /v2beta/stable-image/generate/core endpoint - Core being the company's fast general-purpose hosted model - and returns a finished IMAGE tensor.

The honest framing up front: this whole pack is a paid API wrapper, so you're comparing it against free local generation, and free usually wins for volume. Where Core makes sense: you're on a laptop with no GPU, you want to test ideas in seconds without downloading a model, or you want a quick batch of aspect-ratio variations. The node handles all the API plumbing - upload, auth, decoding the response back into a ComfyUI tensor - so the workflow is just prompt → node → Save Image.

How it works

You provide the prompt and choices; the node POSTs them to Stability as multipart form data and converts whatever comes back (image bytes or base64) into a standard IMAGE tensor. That's the entire mechanism, which is the point - there's nothing to tune. Core is the value pick of the generation trio: cheaper and faster than Ultra, simpler than the SD3 node, good enough composition for most ideation.

Inputs and outputs

• prompt - required. This is a text-to-image model, so prompt it like one: subject, lighting, style, composition. No LoRA, no embeddings, no clip skip - just the words.
• aspect_ratio - required, nine ratios (1:1 default, 16:9, 21:9, 2:3, 3:2, 4:5, 5:4, 9:16, 9:21).
• negative_prompt - optional.
• seed - set it for reproducible results; the range here goes up to 64-bit, so don't be surprised by big numbers.
• style_preset - 18 named looks (anime, cinematic, photographic…), none by default.
• output_format - png / jpeg / webp.
• api_key - per-node override.

Output: a single IMAGE, into Save Image / Preview Image.

Installing it

Part of the aicu-comfyui-stability-ai-api pack:

cd ComfyUI/custom_nodes
git clone https://github.com/aicuai/aicu-comfyui-stability-ai-api.git
cd aicu-comfyui-stability-ai-api
pip install -r requirements.txt

Or ComfyUI Manager → search "aicu-comfyui-stability-ai-api" → install → restart. You'll need a paid Stability AI key from platform.stability.ai: the node's api_key field, the STABILITY_API_KEY env var, or the auto-generated config.ini in the pack folder.

Gotchas

The api_key error is the one you'll actually hit first: if you haven't set a key anywhere and run the node, it fails with "API key must be provided" - that's the pack telling you it looked in the node, the STABILITY_API_KEY env var, and config.ini and found nothing but the placeholder. Also expect the occasional 403 if your prompt trips Stability's moderation - that's a platform policy thing, not a bug, and it's the standing reason some people never adopt this whole pack. And because every image costs credits, do your exploring here before graduating to Ultra. Core is where "just generate me an image" gets answered with the least ceremony.