Inference Providers

by Hugging Face@huggingfaceUnited States and FranceNo ratings yet

Model inference through multiple third-party providers, listed alongside pricing and performance metrics.

Tool type: Model inference routing and comparison catalog
Made for: Developers and teams choosing hosted model inference
Runs on: Hosted model APIs; Requests are served by the listed providers; the published information do not name the specific infrastructure beyond provider names.
Last checked: 28 Oct 2026; 2 product sources

What it does

Inference Providers lets you take an open model from the Hugging Face catalog and send requests to it through a hosted provider. The supported models page lists each model alongside the providers serving it, with input and output prices per million tokens, context length, latency, throughput, and whether tool calling or structured output is supported.

In practice, the table works as a comparison tool. You can look up a model, then weigh cost, speed, and feature support across providers before choosing one. A single model can be offered by several providers, each with its own price and performance profile, so the provider choice stays visible rather than fixed.

It suits developers and small teams who want to test or deploy open models without running their own hardware. Access is through the Hugging Face platform, and you start by signing up for an account. Token pricing is billed separately for input and output and varies widely by model and provider, so check current rates on the supported models page. For free-tier limits or team and enterprise terms, request a quote or consult Hugging Face directly.

What it can do

Lists hosted open models with the providers that serve each one
Shows input and output prices per million tokens for each provider
Shows context length, latency, and throughput per provider and model
Indicates whether a given provider model supports tool calling
Indicates whether a given provider model supports structured output
Lets users compare providers for the same model side by side

Example workflows

Choosing a provider for a model (illustrative): Input: a target open model and a priority such as lowest cost or fastest throughput. Steps: find the model in the supported models table, compare the listed providers on input and output price, latency, and throughput, and check whether tool calling or structured output is needed. Output: a shortlist of providers that fit the chosen priority.
Estimating per-request cost (illustrative): Input: an expected average prompt length and response length for a workload. Steps: read the input and output per-million-token rates for the chosen provider and model, multiply each by the expected token count, and add the two results. Output: an estimated cost per request to use in budgeting.

Before you choose

Provider availability, pricing, and feature support vary by model, so one model's providers may differ from another's.
Prices are quoted per million tokens, with input and output billed separately, so include both in cost estimates.
Some provider rows in the table are incomplete, so confirm the chosen provider's current figures before relying on them.

Listed under

Get started

Create a Hugging Face account, then follow the provider documentation for authentication and billing.

Create or sign in to a Hugging Face account using the sign-up link on the supported models page.
Open the supported models page and find the model you plan to use.
Compare the listed providers for that model on price, latency, throughput, and tool-calling or structured-output support, then select one.

Plans and prices

Prices in USD

Provider rates vary by model and are charged per million input and output tokens. PRO is optional at $9/month and includes $2 in monthly inference credits; it is not a prerequisite for API usage.

PRO$9/moMonthly subscription for an individual account.

$2 of included inference credits monthly
8x ZeroGPU quota and highest queue priority

Inference Providers usageUsage-basedNamed meter: input tokens; see the supported models page for each provider's rate.Named meter: output tokens; see the supported models page for each provider's rate.

Per-token billing for provider models
Rates vary by model and provider

Source: huggingface.co/pricing, checked 28 Oct 2026. Prices can change; check before you buy.

Sources for this overview

Product pages, documentation, and pricing checked 28 Oct 2026. Each source supports part of the overview; availability and prices can change.

Hugging Face Inference Providers · Supported Models ↗huggingface.co
Pricing - Hugging Face ↗huggingface.co

Ratings and reviews

No ratings for Inference Providers yet.

Be the first to rate it

What’s changed

Free account

See what’s changed at Inference Providers

Launches, price moves and pivots, each with a date and a source. Free with an account.

Signals

Free account

See Inference Providers’s signals

Launch traction, hiring and release activity, tracked over time. Free with an account.

Data and security

Free account

See how Inference Providers handles your data

Training on your files, compliance, hosting and the models behind it. Free with an account.

Compare with similar tools

Free account

Compare Inference Providers side by side

Prices, platforms and features against similar tools, with sources. Free with an account.

Where this comes from

How we research

Company website: huggingface.co: product, pricing and policy pages
YC directory mirror: Batch and company name
Web search: Confirms the official site
Social accounts: Accounts the company links, and its posts on X
Human review: Checked by a person
knowledge_base

Last checked 28 Oct 2026. Every field keeps its source and date; unknown means we couldn’t confirm it.