Models API¶
The Models API provides access to a list of models available through the AI Refinery platform. This API allows you to retrieve information about the models you can use for various tasks, such as text generation, completion, and more.
Getting the Supported Model List¶
AsyncAIRefinery.models.list()
¶
Parameters:¶
This method does not require any parameters.
Returns:¶
A list of model objects available for use through AI Refinery. Each model object contains basic information about the model. An example model object can be seen below:
Model(
id='example-model',
created=1234567890,
object='model',
owned_by='AIRefinery',
root='example-model',
parent=None,
permission=[
ModelPermission(
id='example-permission-id',
object='model_permission',
created=1234567890,
allow_create_engine=False,
allow_sampling=True,
allow_logprobs=True,
allow_search_indices=True,
allow_view=True,
allow_fine_tuning=False,
organization='Example Organization',
group=None,
is_blocking=False
)
]
)
Example Usage¶
The Models API lets you list available models via the AsyncAIRefinery
client, enabling you to access models for tasks like text generation and completion.
import asyncio
import os
from air import AsyncAIRefinery, login
from dotenv import load_dotenv
load_dotenv() # loads your ACCOUNT and API_KEY from a .env file
# Authenticate using account and API key retrieved from environment variables
auth = login(
account=str(os.getenv("ACCOUNT")),
api_key=str(os.getenv("API_KEY")),
)
async def list_models():
# Initialize the AI client with authentication details
client = AsyncAIRefinery(**auth.openai())
# Retrieve a list of all available models
models = await client.models.list()
# Print details of each model
for model in models:
print(model)
# Example call to list_models function
if __name__ == "__main__":
asyncio.run(list_models())