Skip to main content
Hugging Face is a platform for building, training, and deploying open-source models. You can use SambaNova as an inference provider to access our models through Hugging Face.

​
Prerequisites

Before starting, ensure you have:

​
Setup

1

Enable SambaNova as an inference provider

Go to your Hugging Face account settings, navigate to Inference Providers, and enable SambaNova.
Hugging Face inference providers settings with SambaNova enabled
2

Add your SambaCloud API key

Enter your SambaCloud API key in the provider configuration.Get your API key from the SambaCloud portal.
Hugging Face API key configuration for SambaNova

​
Usage

There are two ways to use Inference APIs:
  • Custom key: Your requests are sent directly to SambaNova using your own API key. Usage is billed to your SambaNova account.
  • Routed by HF: Your requests are routed through Hugging Face, so you don’t need a separate provider API key. Usage fees are billed to your Hugging Face account.

​
Using the Python SDK

The example below demonstrates how to run a model through SambaNova as the inference provider. You can authenticate with your SambaCloud API key. Make sure you have huggingface_hub version 0.28.0 or newer installed. 
from huggingface_hub import InferenceClient

client = InferenceClient(
    provider="sambanova",
    api_key="your-sambanova-api-key"
)

messages = [
    {
        "role": "user",
        "content": "What is the capital of France?"
    }
]

completion = client.chat.completions.create(
    model="Meta-Llama-3.3-70B-Instruct",
    messages=messages,
    max_tokens=500
)

print(completion.choices[0].message)

​
Additional resources