Baseten Integration with Hugging Face Inference Providers

Hugging Face has integrated Baseten as a supported Inference Provider on the Hugging Face Hub. This integration allows developers to access Baseten's serverless AI infrastructure and a catalog of frontier models directly through Hugging Face's website UI and client SDKs.

Serverless Inference Capabilities

Baseten provides serverless AI infrastructure for training and inference. Through this integration, Baseten initially supports conversational and text-generation tasks on the Hugging Face Hub. This enables immediate access to several popular open-weight Large Language Models (LLMs), including:

DeepSeek V4 Flash
Kimi K3
GLM-5.2

Support for additional model types and tasks beyond text generation is expected to roll out in the future.

Integration and Workflow

Users can access Baseten-hosted models through two primary interfaces: the Hugging Face website UI and official client SDKs.

Website UI Configuration

In the user account settings, Hugging Face users can manage their inference preferences:

API Key Management: Users can set their own Baseten API keys. If no custom key is provided, requests are routed through Hugging Face.
Provider Preference: Users can order providers by preference, which dictates the order they appear in model page widgets and code snippets.

Client SDKs and Agent Harnesses

Baseten is available via the huggingface_hub Python library (version 1.26.1 or higher) and the @huggingface/inference JavaScript library.

Because Hugging Face Inference Providers are integrated into various agent harnesses—such as Pi, OpenCode, Hermes Agents, and OpenClaw—Baseten-hosted models can be plugged into these tools without additional glue code.

Routing and Billing Models

There are two distinct modes for calling Inference Providers like Baseten:

Custom Key Mode: Requests are sent directly to the inference provider using the user's own provider API key. In this mode, the user is billed directly by Baseten.
Routed by HF Mode: Requests are authenticated via a Hugging Face token and routed to Baseten. In this mode, charges are applied to the Hugging Face account at standard provider API rates with no additional markup from Hugging Face.

Credit System for PRO Users

Hugging Face PRO users receive $2 worth of Inference credits every month, which can be used across different providers. Signed-in free users also have access to a small quota of free inference.