Explore the PixAI Model Market, a vibrant hub where creativity meets innovation!
Here, you can discover a diverse array of models and Loras that spark inspiration and ignite your imagination. The market features a handy filter that allows you to easily navigate through options, helping you find exactly what you need—be it character Loras, clothing Loras, or SDXL models.
Search Model/LoRA
You may search Models/LoRAs in the following ways:
Type keywords in the search bar to search for the Model/LoRA you want.
Choose a tag under the search bar first, then type keywords
By using filters, you can search for specific models that meet your needs based on the following criteria:
Base Model: Choose from a range of models, from SD1.5 to Realistic, with all base models supported for LoRA training.
Source: Filter between PixAI-trained models and external sources.
Categories: Select from various categories such as Character, Style, Pose, Detail Enhancement, and more.
Submission Time: Narrow your search by filtering LoRAs submitted within specific time frames, such as the past 7 days or the past 30 days.
Use Model/LoRA
By clicking on your preferred model or LoRA, you'll be redirected to the model detail page. There, simply click on "Use this model/LoRA" in the top right corner to bookmark it.
This will take you to the generation page, where the selected model will be set as the default for your generation settings, or the LoRA will be added to your generation task.
Train Your Own LoRA
Didn’t find a suitable LoRA? You can train your own!
Follow these simple steps, and you’ll have your LoRA ready in just a few minutes.
Upload up to 100 images. The more images you provide, the higher the quality of your LoRA. Make sure the images focus on a single subject or character and are at least 768 pixels in size for optimal results. Additionally, please remove any watermarks manually to enhance the quality of your training data.
Name your LoRA and select a base model. It’s advisable to choose the model you use most frequently as your base model to achieve the best results. This ensures compatibility and enhances the performance of your LoRA.
Enter some trigger words for your LoRA. These will appear in the prompt box after the LoRA is added. Keep in mind that you cannot change these trigger words later, so choose wisely. If you make a mistake, you’ll need to train a new LoRA.
Once the training is complete, which takes just a few minutes, you’ll receive a system notification.
You’ll then be able to use and publish your LoRA. Be sure to update the category of your LoRA so others can easily find it!
Click here for more details about LoRA training on PixAI!
Upload Models/LoRAs
By clicking "Upload," you can share your own models and LoRAs with the PixAI community.
Follow these steps to upload your model/lora:
Please provide the details for your model, including the name, type, tags, and description. We highly recommend tagging your model accurately to enhance its visibility and exposure.
For models, please also add the quality tag here.
For LoRAs, please make sure to select the category.
2. Please specify the version for your model or LoRA, and fill out the trigger words if applicable.
3. Please upload a cover image and submit the file of your model/LoRA (Only ,cptk, .pt, .safetensors files are allowed)
