Fofr: Latent Consistency Model
fofr/latent-consistency-model
Access Latent Consistency Model from Fofr using the Puter.js AI API.
Add to your appModel Card
Latent Consistency Model is an implementation of LCM, a distillation technique that lets a diffusion model generate images in as few as 1 to 4 steps instead of the usual 20 to 50, cutting inference time sharply.
This port is based on Dreamshaper v7, the same Stable Diffusion 1.5 fine-tune the original LCM paper distilled its released checkpoint from. Replicate lists a runtime of about 0.6 seconds per image on an A100 GPU.
It supports img2img, large batch generation, and canny ControlNet conditioning. The reduced step count trades some fine detail for speed, making it suited to real-time previews, rapid prototyping, and workflows where turnaround matters more than maximum fidelity.
Cost Per Image $0.0014
per generation
Configuration second
resolution
Release Date N/A
Add Latent Consistency Model to your app
Use Latent Consistency Model in your app with the Puter.js AI API, no API keys or setup required.
// npm install @heyputer/puter.js
import { puter } from '@heyputer/puter.js';
puter.ai.txt2img("A serene mountain landscape at sunset", {
model: "fofr/latent-consistency-model"
}).then(image => {
document.body.appendChild(image);
});
<html>
<body>
<script src="https://js.puter.com/v2/"></script>
<script>
puter.ai.txt2img("A serene mountain landscape at sunset", {
model: "fofr/latent-consistency-model"
}).then(image => {
document.body.appendChild(image);
});
</script>
</body>
</html>
More AI Models From Fofr
SDXL Emoji
SDXL Emoji is a fine-tune of Stable Diffusion XL trained on Apple's emoji set, producing new emoji-style images from text prompts. It was trained with Dreambooth LoRA combined with textual inversion (pivotal tuning), a method that pairs a LoRA-trained concept with a newly learned token. Prompts need the trigger word TOK to activate the emoji style, for example "TOK emoji of a cat wearing sunglasses". Output keeps the glossy, rounded, flat-colored look of Apple's emoji library rather than general SDXL photorealism. It suits chat apps, custom emoji packs, and other cases that call for a recognizable emoji look instead of a generic illustration.
ImageSDXL Multi ControlNet LoRA
SDXL Multi ControlNet LoRA is an SDXL image generation endpoint that accepts up to three ControlNet conditioning inputs at once, alongside custom LoRA weights loaded from Replicate. The available ControlNet types include canny edge detection, depth (both Midas and LeReS), soft edge (HED and PiDiNet), OpenPose pose estimation, QR Monster for illusion-style images, and lineart in standard and anime variants. Combining several of these, such as a depth map and a canny edge map together, gives more precise control over structure and pose than a single ControlNet alone. It also supports img2img and inpainting on top of the ControlNet inputs, plus an SDXL refiner pass. This suits use cases that need tight structural control paired with a custom trained style, such as product shots in a fixed pose or branded illustration in a specific LoRA style.
ImageSticker Maker
Sticker Maker generates die-cut sticker-style images from text prompts, with bold outlines, flat colors, and a transparent background. It builds on AlbedobaseXL, an SDXL fine-tune, combined with a LoRA trained specifically for sticker output (ArtificialGuyBr's stickers LoRA) and LayerDiffuse for the transparency. Predictions typically complete in about 5 seconds on Replicate. The output is ready to use directly as a sticker asset without a separate background removal step, useful for chat stickers, print-on-demand designs, and other cases that call for a clean cutout look rather than a full photographic scene.
Frequently Asked Questions
You can access Latent Consistency Model by Fofr through Puter.js AI API. Include the library in your web app or Node.js project and start making calls with just a few lines of JavaScript — no backend and no configuration required.
Latent Consistency Model is free to integrate using the Puter.js AI API. With the User-Pays Model, you can add AI to your app for $0, since users cover their own AI usage through their Puter account.
| Price | |
|---|---|
| Per image | $0.0014 |
Yes — the Latent Consistency Model API works with any JavaScript framework, Node.js, or plain HTML through Puter.js. Just include the library and start building. See the documentation for more details.
Add Latent Consistency Model to your app for free
Developers can integrate Latent Consistency Model for free using the Puter.js AI API.
With the User-Pays Model, each user covers their own AI usage instead of the developer.