Supergateway hosting
Put a stdio MCP server on the network without rewriting it.
Supergateway hosting at AIHostBox means a dedicated, single-tenant Supergateway instance on our infrastructure: we install it, pin it to a reviewed release, put it behind TLS on your domain, back it up nightly and keep it patched. Plans start at $5.99 per month and the admin account, the data and any API keys are yours.
Supergateway is a small open-source bridge from Supercorp that solves one MCP problem: most MCP servers speak stdio, which only works when the client launches them as a local process. Supergateway starts that process itself and republishes it over the network as Streamable HTTP, Server-Sent Events or WebSockets. It also works the other way round, turning a remote SSE or Streamable HTTP server into stdio for clients that only support local tools.
It runs from npx or from the supercorp/supergateway image, which has variants with uvx for Python-based servers and Deno for TypeScript ones. The whole point is remote reachability, so a laptop defeats it: the tool disappears when the laptop sleeps, and a cloud agent cannot reach a home connection anyway. On a server the MCP endpoint has a fixed HTTPS address that Claude, a Letta agent or a Dify workflow can call at any time.
Supergateway hosting plans
Every plan is a dedicated single-tenant instance with NVMe storage, nightly off-box backups, TLS and managed patching. Quarterly, semi-annual and annual terms are discounted at checkout. Model tokens are not included — you bring your own key.
How we run Supergateway
The supercorp/supergateway container (with the uvx or Deno variant where the tool needs it) on port 8000, launching your stdio MCP server as a child process and serving it over Streamable HTTP or SSE.
What we configure for you
- --stdio command for your MCP server and any secrets it needs as environment variables
- --outputTransport streamableHttp (or sse / ws) with the matching --streamableHttpPath or --ssePath
- --stateful and --sessionTimeout when the wrapped server keeps per-session state
- --healthEndpoint for uptime checks and --cors limited to the origins you name
- TLS on your subdomain and an access check in front, since the gateway has no inbound auth
Which plan fits
Starter (2 GB) runs one or a few lightweight MCP servers, which is what most people need. Plus (4 GB) suits several gateways side by side or servers that spawn heavier runtimes. Pro (8 GB) is for tools that do real work themselves, such as headless browsers or large repository indexing.
The honest limits
The gateway adds no authentication, rate limiting or tool filtering of its own. For central policy across many MCP servers, a fuller MCP gateway is the better fit.
What people run Supergateway for
Remote access to a local tool
Keep the server code as it is.
Bridge for hosted assistants
Make an existing tool reachable from anywhere.
Python MCP servers without packaging
Run a uvx-installed server such as mcp-server-git and publish it over HTTP.
One tool, several agents
Point different agent frameworks at the same MCP endpoint instead of installing the tool in each.
Licence and version policy
| Item | Position |
|---|---|
| Upstream licence | MIT |
| Sold as managed hosting | Yes |
| Tenancy | Single tenant — one customer per instance |
| Patching | Critical advisories within 24 hours, high severity within 72 |
Our full position on licences, including the applications we decline to sell, is on the licensing page.
Reading on Supergateway
Works well with Supergateway
Supergateway hosting: your questions
Which transport should my client use?
Can one instance host several MCP servers?
How are requests authenticated?
Which MCP servers can it run?
Does Supergateway store any data?
Run Supergateway without running a server
A dedicated Supergateway instance, deployed and maintained by us. Month to month, cancel whenever.