Models
Overview
Models are what power the AI part of your workflows. In Floyo, you can use any model supported through ComfyUI, including both open-source and closed/API-based models.
Overview
Model Type | Access Method | Examples |
|---|---|---|
Open Source | Pre-loaded or community models | Stable Diffusion, FLUX, SDXL |
Custom LoRAs | Upload your own | Personal fine-tunes, style LoRAs |
API Models | Floyo Partner Nodes | Kling, Stability, Luma, OpenAI |
Open Source Models
Floyo supports all the major open source models you're used to:
Category | Description |
|---|---|
Text2Image models | Generate images from text prompts |
Image2Image models | Transform existing images |
AI video models | Generate and edit video content |
LLMs | Large language models for text generation |
Inpainting and upscaling models | Edit and enhance images |
LoRAs and embeddings | Style and concept fine-tunes |
Any model that works in ComfyUI can work in Floyo. Your models show up in the interface and can be used in any workflow.
Adding Models
Upload LoRA Models
You can upload your own LoRA models directly to your account:
- Open the File Browser in the left sidebar
- Navigate to the models folder
- Upload your LoRA files
See Upload LoRA Models for detailed instructions.
Request Community Models
For checkpoints, embeddings, and other model types:
- Contact Floyo to request the model be added
- Provide the model name and source (e.g., Hugging Face, CivitAI)
- The team will add it to the community models
See Request Community Model Upload for the request process.
Closed-source Models via API
Floyo also supports running closed-source or third-party models using API calls using the Floyo Partner nodes
This gives you full flexibility:
- Combine open models with closed ones in the same workflow
- Use external tools mid-process
- Trigger outputs based on responses from other systems
See Floyo Partner Nodes for available API integrations.
Model Privacy and Control
Your models stay private unless you choose to share them.
Context | Privacy |
|---|---|
Personal account | Your LoRAs are private to you |
Team account | Admins control which models are available to the team |
Enterprise | Full control over model access and usage |
In team environments, you can manage access, control usage, and define what models are available in which workspace.
Model Locations in File Browser
Folder | Contents |
|---|---|
models | Your uploaded custom models (LoRAs) |
community_models | Shared models available to all users |
Tips & Best Practices
- Start with community models: Explore what's available before uploading custom models
- Organize LoRAs: Use descriptive filenames for easy identification
- Check compatibility: Ensure models match your workflow requirements (e.g., SD 1.5 vs SDXL)
- Monitor storage: Models count against your storage allocation
Troubleshooting
Model not appearing in loader node?
Solution: Refresh the canvas or check that the model uploaded successfully in the File Browser. Some loader nodes may need to be re-added to detect new models.
Upload failed?
Solution: Check that your file is a supported format and within size limits. LoRA files should be .safetensors or .pt format.
Model produces unexpected results?
Solution: Verify the model is compatible with your workflow. For example, SDXL LoRAs won't work with SD 1.5 checkpoints.