Inference Providers documentation
Pricing and Billing
Pricing and Billing
Access 200+ models from leading AI inference providers with centralized, transparent, pay-as-you-go pricing. No infrastructure management required—just pay for what you use, with no markup from Hugging Face.
Included Monthly Credits
PRO users and Team or Enterprise organizations receive monthly credits to use with Inference Providers. Free users can use Inference Providers by purchasing credits, or by upgrading to PRO:
| Account Type | Monthly Credits | Can be spent on | Extra usage (pay-as-you-go) |
|---|---|---|---|
| Free Users | None | - | yes (credits purchase required) |
| PRO Users | $2.00 | All Hugging Face compute services | yes |
| Team or Enterprise Organizations | $2.00 per seat | All Hugging Face compute services | yes |
Credits included with a PRO, Team, or Enterprise subscription are general-purpose compute credits: on top of Inference Providers, they can also be spent on Inference Endpoints, upgraded CPU & GPU hardware for Spaces (including ZeroGPU usage beyond your quota), and Jobs. They are credited every month and applied automatically before any pay-as-you-go usage is billed.
Your monthly credits automatically apply when you route requests through Hugging Face. For Team or Enterprise organizations, credits are shared among all members.
Pay-as-you-Go Details
To benefit from Team or Enterprise included credits, you need to explicitly specify the organization to be billed when performing the inference requests. See the Organization Billing section below for more details.
All users can use the API beyond their included monthly credits (if any) by purchasing additional credits. This ensures uninterrupted access to models for production workloads.
Hugging Face charges you the same rates as the provider, with no additional fees. We just pass through the provider costs directly.
You can track your spending anytime on your billing page.
Inference Providers Usage Breakdown
View detailed usage information for Inference Providers directly from your settings. Visit your Inference Providers Settings to see your usage for the past month, broken down by model and provider.
The same detailed view is available for organizations subscribed to a paid plan under the organization’s settings.

Retrieve organization usage via the API
This feature is part of the Enterprise plan and above.
Organizations can also retrieve their Inference Providers usage programmatically using the dedicated API endpoint. It returns a daily time series of usage broken down by member, model, and provider, including the number of requests, the cost (in cents), and input, output, and cached input token counts (token counts are only reported for chat completions). Costs are rounded to the nearest cent for each row, so very small amounts may show as 0 and the sum of rows may differ slightly from your invoice. Requests appear in the results up to 2 hours after they are made.
Use a token with permission to read the organization’s billing to call it:
curl "https://huggingface.co/api/organizations/my-org-name/billing/usage/inference?startDate=2026-09-01T00:00:00Z&endDate=2026-09-30T00:00:00Z&limit=31" \
-H "Authorization: Bearer $HF_TOKEN"The startDate and endDate query parameters set the time window: it defaults to the current month and must fall within the last 12 months. Results are paginated by day: use the limit query parameter to set how many days are returned per page (7 by default, up to 31), and follow the rel="next" URL in the Link response header to fetch the next page.
HF-Inference cost
As you may have noticed, you can select to work with "hf-inference" provider. This service used to be called “Inference API (serverless)” prior to Inference Providers. From a user point of view, working with HF Inference is the same as with any other provider. Past your included credits (if any), you get charged for every inference request based on the compute time x price of the underlying hardware.
For instance, a request to black-forest-labs/FLUX.1-dev that takes 10 seconds to complete on a GPU machine that costs $0.00012 per second to run, will be billed $0.0012.
As of July 2025, hf-inference focuses mostly on CPU inference (e.g. embedding, text-ranking, text-classification, or smaller LLMs that have historical importance like BERT or GPT-2).
Billing for Team and Enterprise organizations
For Team & Enterprise organizations, it is possible to centralize billing for all of your users. Each user still uses their own User Access Token but the requests are billed to your organization. This can be done by passing "X-HF-Bill-To: my-org-name" as a header in your HTTP requests.
Enterprise organizations with Resource Groups set up can also attribute inference costs to a specific resource group by passing the resource group’s ID instead of the org name: "X-HF-Bill-To: <resource-group-id>". The user’s token must be a member of the resource group.
Team & Enterprise organizations receive a pool of free usage credits based on the number of seats in the subscription. Inference Providers usage can be tracked on the organization’s billing page. Team & Enterprise organization administrators can also set a spending limit and disable a set of Inference Providers from the organization’s settings.
When a request is billed to an organization, the providers enabled or disabled in the organization’s settings apply instead of the requesting user’s personal settings. For example, if an organization disables a provider, requests billed to that organization will not be routed to it, even if the user has it enabled in their own settings. Explicitly requesting a provider that is disabled by the organization returns an error.

To bill your organization, use the bill_to parameter when initializing the client.
from huggingface_hub import InferenceClient
client = InferenceClient(bill_to="my-org-name")
completion = client.chat.completions.create(
model="deepseek-ai/DeepSeek-V3-0324",
messages=[
{
"role": "user",
"content": "How many 'G's in 'huggingface'?"
}
],
)
print(completion.choices[0].message)To bill your organization when using OpenAI’s Python client, set the X-HF-Bill-To header using extra_headers on the completions.create method call.
import os
from openai import OpenAI
client = OpenAI(
base_url="/static-proxy?url=https%3A%2F%2Frouter.huggingface.co%2Fv1%26quot%3B%3C%2Fspan%3E%2C
api_key=os.environ["HF_TOKEN"],
)
completion = client.chat.completions.create(
model="deepseek-ai/DeepSeek-V3-0324",
messages=[
{
"role": "user",
"content": "How many 'G's in 'huggingface'?"
}
],
extra_headers={"X-HF-Bill-To": "my-org-name"},
)
print(completion.choices[0].message)To bill your organization when making direct HTTP requests, include the X-HF-Bill-To header.
import os
import requests
API_URL = "/static-proxy?url=https%3A%2F%2Frouter.huggingface.co%2Fv1%2Fchat%2Fcompletions%26quot%3B%3C%2Fspan%3E
headers = {"Authorization": f"Bearer {os.environ['HF_TOKEN']}", "X-HF-Bill-To": "my-org-name"}
payload = {
"messages": [
{
"role": "user",
"content": "How many 'G's in 'huggingface'?"
}
],
"model": "deepseek-ai/DeepSeek-V3-0324",
}
response = requests.post(API_URL, headers=headers, json=payload)
print(response.json()["choices"][0]["message"])Similarly in JavaScript:
If you are using the JavaScript InferenceClient, you can set the billTo attribute at a client level to bill your organization.
import { InferenceClient } from "@huggingface/inference";
const client = new InferenceClient(process.env.HF_TOKEN, { billTo: "my-org-name" });
const completion = await client.chat.completions.create({
model: "deepseek-ai/DeepSeek-V3-0324",
messages: [
{
role: "user",
content: "How many 'G's in 'huggingface'?",
},
],
});
console.log(completion.choices[0].message.content);To bill your organization with the OpenAI JavaScript client, set the X-HF-Bill-To header using the defaultHeaders option on the completions.create method call.
import OpenAI from "openai";
const client = new OpenAI({
baseURL: "/static-proxy?url=https%3A%2F%2Frouter.huggingface.co%2Fv1%26quot%3B%3C%2Fspan%3E%2C
apiKey: process.env.HF_TOKEN,
});
const completion = await client.chat.completions.create(
{
model: "deepseek-ai/DeepSeek-V3-0324",
messages: [
{
role: "user",
content: "How many 'G's in 'huggingface'?",
},
],
},
{
headers: {
"X-HF-Bill-To": "my-org-name",
},
}
);
console.log(completion.choices[0].message.content);When using fetch, include the X-HF-Bill-To header to bill your organization.
import fetch from "node-fetch";
const response = await fetch(
"/static-proxy?url=https%3A%2F%2Frouter.huggingface.co%2Fv1%2Fchat%2Fcompletions%26quot%3B%3C%2Fspan%3E%2C
{
method: "POST",
headers: {
Authorization: `Bearer ${process.env.HF_TOKEN}`,
"Content-Type": "application/json",
"X-HF-Bill-To": "my-org-name",
},
body: JSON.stringify({
model: "deepseek-ai/DeepSeek-V3-0324",
messages: [
{
role: "user",
content: "How many 'G's in 'huggingface'?",
},
],
}),
}
);
console.log(await response.json());