---
title: "Baseten on Hugging Face Inference Providers 🔥"
description: "Baseten is now a supported Inference Provider on the Hugging Face Hub, extending serverless inference on model pages and through the Hugging Face Python and JavaScript SDKs. As part of the initial integration, it supports conversational and text-generation tasks, including access to open-weight models such as Kimi K3, latest DeepSeek V4 Flash, and GLM-5.2, with additional tasks planned to roll out soon. Developers can use a custom provider key for direct calls or authenticate with a Hugging Face token so requests are routed to Baseten; model pages and account settings support provider compatibility and preference ordering. Python and JavaScript OpenAI-compatible examples use the Hugging Face router, while integrated agent harnesses can access Baseten-hosted models without extra glue code."
---

# Baseten on Hugging Face Inference Providers 🔥

[Hugging Face](https://yomu.fyi/company/hugging-face) · Alex Ker, Roland Crosby, Sid Shanker, Johan, Célina Hanouti, Simon Brandeis, Lucain Pouget, merve · Aug 6, 2026

**Type:** Announcement

## Summary

Baseten is now a supported Inference Provider on the Hugging Face Hub, extending serverless inference on model pages and through the Hugging Face Python and JavaScript SDKs. As part of the initial integration, it supports conversational and text-generation tasks, including access to open-weight models such as Kimi K3, latest DeepSeek V4 Flash, and GLM-5.2, with additional tasks planned to roll out soon. Developers can use a custom provider key for direct calls or authenticate with a Hugging Face token so requests are routed to Baseten; model pages and account settings support provider compatibility and preference ordering. Python and JavaScript OpenAI-compatible examples use the Hugging Face router, while integrated agent harnesses can access Baseten-hosted models without extra glue code.

## Context

Baseten joins Hugging Face's Inference Provider ecosystem to broaden serverless inference available directly on model pages, client SDKs, and integrated agent tools. The initial integration focuses on conversational and text-generation tasks, with additional task support planned.

## Approach / What changed

Hugging Face exposes Baseten through model-page provider selection, custom-key or Hugging Face-routed requests, the huggingface\_hub Python SDK, the @huggingface/inference JavaScript SDK, OpenAI-compatible router calls, and integrations with agent harnesses. Direct requests use a provider key, while routed requests use a Hugging Face token and charge the Hugging Face account at standard provider API rates.

## Takeaways

- The initial Baseten integration supports conversational and text-generation tasks and provides access to models including Kimi K3, latest DeepSeek V4 Flash, and GLM-5.2.
- Requests can use a custom Baseten key for direct provider calls or be routed by Hugging Face with a Hugging Face token; direct and routed requests are billed through the corresponding account.
- Baseten is available through the Hugging Face Python and JavaScript SDKs, OpenAI-compatible router examples, and several integrated agent harnesses without extra glue code.

**Tags:** [LLMs](https://yomu.fyi/topic/llm), [Open Source](https://yomu.fyi/topic/open-source), [Python](https://yomu.fyi/topic/python), [Serverless](https://yomu.fyi/topic/serverless)

- Source: [Hugging Face](https://huggingface.co/blog/baseten)
- Source URL: https://huggingface.co/blog/baseten
- Ingested by Yomu: 2026-08-29T05:40:46.982Z

[Read original post](https://huggingface.co/blog/baseten)