MCP Gateway · Claude
Connect Hugging Face to Claude
Analysts, researchers and product managers often need to know what is new in machine learning without reading every release. Added to Claude as a custom connector through the Speakeasy MCP Gateway, Hugging Face lets people search papers and models in a conversation, and run Space tasks like OCR on their own files.
- Category
- AI & search
- Tools
- 10
- Authentication
- OAuth
- Claude
- Web, desktop and mobile
Behind the MCP Gateway
Run Hugging Face MCP behind the MCP Gateway
One URL for every agent, sign-in through your IdP, RBAC down to the tool, and every Hugging Face call inspected and logged.
Set up
Add Hugging Face to Claude in 4 steps
- 01
Add Hugging Face from the catalog
An admin finds Hugging Face in the MCP Catalog (MCP Gateway > MCP > Catalog) and clicks Add. Speakeasy registers the vendor's server as a remote MCP server and proxies sessions to it. Nothing is copied or re-hosted.
Hugging Face uses OAuth. Servers that support dynamic client registration get their OAuth configuration applied automatically.
Where MCP-scoped policies are enabled, the install offers a recommended risk policy for the new server that detects secrets and PII and inspects tool requests and responses.
- 02
Choose who can use Hugging Face
Give teams the
mcp:connectscope on the Hugging Face server through team access or a role. The organization gateway works out each person's servers from their live access, so granting access adds Hugging Face to their gateway on the next call, and removing it takes it away. - 03
Connect Claude
Add the gateway URL as a custom connector, then connect it from Customize > Connectors.
On Team and Enterprise plans, an Owner opens Organization settings > Connectors, chooses Add > Custom > Web and pastes the gateway URL.
Each person opens Customize > Connectors and clicks Connect to sign in.
On Pro and Max plans, each person adds it themselves under Customize > Connectors, with + Add > Add custom connector.
- 04
Sign in and use Hugging Face from Claude
The first time Claude connects, it opens a browser so the person can sign in to Speakeasy through your identity provider.
Hugging Face needs its own OAuth grant, so each person connects it once. Until they do, Hugging Face calls fail while the rest of the gateway keeps working.
From then on, Hugging Face's tools appear in Claude, limited to what the person's access allows, and every call lands in the gateway's tool logs.
No client ID or secret is needed: the gateway's sign-in handles client registration. Each person only sees the servers their role can reach.
Use cases
What Claude users do with Hugging Face
Summarize recent research on a topic
Ask Claude for recent papers on a topic like speech recognition, with links, and get a plain summary of what each one claims.
Extract text from a scanned page
Claude runs an OCR task through a Hugging Face Space on a scanned document, then tidies the text into something you can quote.
Explain a model before a decision
When a vendor names the model behind their product, ask Claude for its Hub details and an explanation of what the model is meant for.
Tools
10 tools
Each Hugging Face tool can be allowed or denied per team and role at the MCP Gateway before Claude sees it.
hf_whoami
Hugging Face tools are being used by the authenticated user
space_search
Find Hugging Face Spaces using semantic search. IMPORTANT Only MCP Servers can be used with the dynamic_space toolInclude links to the Space when presenting the results.
hub_repo_search
Search Hugging Face repositories with a shared query interface. You can target models, datasets, spaces, or aggregate across multiple repo types in one call. Use space_search for semantic-first discovery of Spaces. Include links to repositories in your response.
paper_search
Find Machine Learning research papers on the Hugging Face hub. Include 'Link to paper' When presenting the results. Consider whether tabulating results matches user intent.
hub_repo_details
Get details for one or more Hugging Face repos (model, dataset, or space). Auto-detects type unless specified.
hf_doc_search
Search and Discover Hugging Face Product and Library documentation. Send an empty query to discover structure and navigation instructions. Knowledge up-to-date as at 8 April 2026. Combine with the Product filter to focus results.
Control
What Speakeasy can see and stop in Claude
The gateway covers every client the same way. The rest depends on what Claude exposes to Speakeasy.
Tool logs
Gateway
Every Hugging Face call is logged with the person and the client that made it.
Access by role
Gateway
People reach only the Hugging Face tools their team or role allows.
Request and response blocking
Gateway
A risk policy scoped to Hugging Face can reject a tool request before it reaches the server, or withhold a result from Claude. MCP-scoped policies are enabled per organization.
Anthropic inference hooks
Claude
On Claude Enterprise, each prompt is sent to Speakeasy before inference and denied prompts are blocked (beta). Inference hooks can't rewrite or redact a prompt.
Conversation import
Claude
Claude conversations can be imported through Anthropic's Compliance API for review after the fact.
- Custom connectors need a Claude plan that supports them.
- Budgets don't cover Claude web and desktop usage.
Questions
Frequently asked questions
Can Hugging Face prompts in Claude be checked before inference?
How do people on Pro and Max plans add Hugging Face to Claude?
How do Claude users get Hugging Face?
Are Hugging Face calls from Claude logged?
Connect more