Baseten Joins Hugging Face Hub as a Supported Inference Provider
Summary
- • Baseten is now a supported Inference Provider on the Hugging Face Hub, enabling serverless model access through HF interfaces
- • Developers can access popular open-weight LLMs — including DeepSeek V4 Flash, Kimi K3, and GLM-5.2 — via Baseten through HF SDKs
- • Integration works with Python and JavaScript HF client SDKs and major agent harnesses including Pi, OpenCode, and Hermes Agents
- • Two billing modes available: direct to Baseten account using a Baseten API key, or routed through HF billed to the HF account
Details
Baseten becomes a Hugging Face Inference Provider
Baseten becomes a supported Inference Provider on the Hugging Face Hub, joining a growing ecosystem of serverless AI inference options available directly from model pages.
Initial launch supports popular open-weight LLMs
Supports Kimi K3, DeepSeek V4 Flash, and GLM-5.2 for conversational and text-generation tasks at launch; additional task types coming soon.
Available via Python and JavaScript HF SDKs
Available via huggingface_hub (>=1.26.1) for Python and @huggingface/inference for JavaScript; also accessible with OpenAI-compatible clients pointing to the HF router URL.
Two billing modes: direct or routed through Hugging Face
Developers using their own Baseten API key are billed directly by Baseten; requests routed through HF authenticate with an HF token and charges apply to the HF account — no Baseten account required.
Compatible with major agent harnesses out of the box
Baseten-hosted models are accessible in agent frameworks including Pi, OpenCode, Hermes Agents, and OpenClaw without additional integration code.
Baseten–Hugging Face integration details as announced by Hugging Face.
What This Means
Baseten joining the Hugging Face Inference Provider ecosystem continues a pattern of AI infrastructure companies integrating with HF to reach developers at their primary workflow hub. For users, it widens the range of frontier open-weight models accessible through a single interface. For Baseten, it's a significant distribution play that reduces sign-up friction. The dual-billing model also signals growing maturity in how inference marketplaces handle routing and monetization.
Sources
- Baseten on Hugging Face Inference Providers 🔥Hugging Face