v3

OpenAPI 3.1.0MIT2026-07-31192456807.8 KB

Models

Upload a custom model or adapter

Upload a custom model or adapter from Hugging Face or S3

post/models

Request body

model_namestring required

The name to give to your uploaded model

model_sourcestring required

The source location of the model (Hugging Face repo or S3 path)

model_type'model' | 'adapter'

Whether the model is a full model or an adapter

hf_tokenstring

Hugging Face token (if uploading from Hugging Face)

descriptionstring

A description of your model

base_modelstring

The base model to use for an adapter if setting it to run against a serverless pool. Only used for model_type adapter.

lora_modelstring

The lora pool to use for an adapter if setting it to run against, say, a dedicated pool. Only used for model_type adapter.

Example request

{
  "model_name": "Qwen2.5-72B-Instruct",
  "model_source": "unsloth/Qwen2.5-72B-Instruct",
  "model_type": "model",
  "hf_token": "hf_examplehuggingfacetoken",
  "description": "Finetuned Qwen2.5-72B-Instruct by Unsloth",
  "base_model": "Qwen/Qwen2.5-72B-Instruct",
  "lora_model": "my_username/Qwen2.5-72B-Instruct-lora"
}

Response

Model / adapter upload job created successfully

messagestring required

Example response

{
  "data": {
    "job_id": "job-a15dad11-8d8e-4007-97c5-a211304de284",
    "model_name": "necolinehubner/Qwen2.5-72B-Instruct",
    "model_id": "model-c0e32dfc-637e-47b2-bf4e-e9b2e58c9da7",
    "model_source": "huggingface"
  },
  "message": "Processing model weights. Job created."
}