Table of contents
You can create your own model on Replicate, and then either push a custom model to it or use it as a destination for fine-tunes.
You can create a new model on the web or via the API.
Visit replicate.com/create to choose a name for your model, and specify whether it should be public or private.
You must also choose the model hardware. This will affect how the model performs and how much it costs to run. Take a look at the pricing docs to see the specifications of the different hardware and how much it costs.
There's also an API endpoint for creating models. You can use it to automate the creation of models, including fine-tunes of other models.
Note that there is a limit of 1,000 models per account. For most purposes, we recommend using a single model and pushing new versions of the model as you make changes to it.
Check out the HTTP API reference for more detailed documentation about the model creation endpoint.
Here's an example that uses cURL to create a model with a given owner, name, visibility, and hardware:
The response is a JSON object of the created model:
To see all the hardware available for your model to run, consult our endpoint for listing hardware.
To compare the price and specifications of these hardware types, check out the pricing page.
Once you've created a model on Replicate, you can push to it with Cog or use it as a training destination for fine-tunes.