HTML.now MCP Server

npx add-mcp 'https://html.now/mcp'

What is the HTML.now MCP server?

It lets an AI assistant use HTML.now converters as structured tools. The client can discover supported routes, start a conversion, check its status, and return a secure download URL.

MCP endpoint

https://html.now/mcp

The server uses stateless Streamable HTTP transport. Your client must support remote HTTP servers and custom authentication headers.

Connect HTML.now to Codex

1. 1 Create an API key Sign in and open Dashboard → API Access.
2. Save the key

   export HTMLNOW_API_KEY="htmlnow_your_key_here"
   

3. Add the remote server

   codex mcp add html-now \
     --url https://html.now/mcp \
     --bearer-token-env-var HTMLNOW_API_KEY
   

4. Check the connection

   codex mcp get html-now
   

Connect other MCP clients

Configuration names vary, but the URL and authentication header are the same:

{
  "mcpServers": {
    "html-now": {
      "url": "https://html.now/mcp",
      "headers": { "X-Api-Key": "htmlnow_your_key_here" }
    }
  }
}

Authentication

Create or regenerate the key in the API dashboard. You may also send it as Authorization: Bearer …. It uses the same credits, limits, and ownership rules as the REST API and is never returned in tool output.

Verify the MCP connection

Ask your client: “List the available HTML.now converters.” This calls list_converters without using credits. Then ask for API usage to confirm authentication.

If it fails: 401 means an invalid key; 402 means insufficient credits; 429 means the request limit was reached.

HTML.now MCP tools

Tool	Purpose	Credits
list_converters	Search valid conversion routes and formats.	No
convert_image_url	Fetch a public URL and start the selected conversion. The legacy name is retained for API compatibility.	Yes
get_conversion_status	Check a job and obtain its temporary download.	No
get_api_usage	Read credits, recent usage, and limits.	No

Example workflow

Ask your assistant:

Use the html-to-pdf converter for the file at this public URL, wait for it to finish, and return the download link.

Security and limits

• Only public HTTP and HTTPS source URLs are accepted.
• Private, loopback, link-local, and reserved addresses are blocked.
• Every redirect is validated before it is followed.
• Source downloads are size-limited and use temporary storage.
• Files follow the normal HTML.now retention policy.

Prefer direct HTTP integration? Read the HTML.now REST API documentation.