Hugging Face Free Monthly Inference Credits

Overview

Hugging Face routes inference requests to a network of partner providers through a single API and token. The official pricing documentation lists monthly credits for every account type, and Hugging Face passes through provider rates with no added fee: "Hugging Face charges you the same rates as the provider, with no additional fees."

Free Tier Quotas & Limits

• Credits are issued every month and applied automatically before any pay-as-you-go usage is billed.
• Free-tier credits can be spent on Inference Providers. PRO, Team, and Enterprise credits also cover Inference Endpoints, upgraded Spaces hardware including ZeroGPU usage, and Jobs.
• Once monthly credits are exhausted, all users can continue by purchasing additional credits; there is no hard stop.
• Credits only apply when requests are routed through Hugging Face. If you set a custom provider key, the provider bills you directly and credits do not apply.
• The hf-inference provider focuses mostly on CPU inference (embeddings, text ranking, classification, smaller LLMs); GPU-accelerated LLM requests route to external providers.

How to Get Started

1. Create a free account at huggingface.co/join.
2. Create a fine-grained access token with the permission to call Inference Providers.
3. Point an OpenAI-compatible client at https://router.huggingface.co/v1, or use the huggingface_hub / @huggingface/inference clients.
4. Track consumption on the billing page, with per-model and per-provider breakdowns in Inference Providers settings.