Helm7
Start free

One platform for the app and everything behind it.

Helm7 is a platform as a service and a backend as a service in one. It builds and runs your application, and it hosts the Postgres, auth, storage, jobs and domains the application needs, on one bill and behind one API.

Start freeRead the quickstart
Platform as a service
A platform as a service (PaaS) builds your code and runs it for you, so you deploy an application rather than manage servers. Heroku made the idea popular; Vercel, Railway and Render are current examples.
In Helm7. Builds from a Git repository or a Dockerfile, runs web services, workers and cron jobs, and attaches domains with certificates.
Backend as a service
A backend as a service (BaaS) hosts the parts behind an application that every app needs: a database, user sign-in, file storage. Supabase and Firebase are the best-known examples.
In Helm7. Postgres 18 with pgvector, managed auth for your users, and S3-compatible storage with no egress charge.
Managed Postgres
Managed Postgres hosting means someone else runs the database server: provisioning, connection pooling, backups and upgrades. You get a connection string.
In Helm7. One database and role per Product, pooled and direct connection strings set for you, daily backups, and point-in-time recovery on Pro.

A Product is the whole application.

Other platforms give you a project and leave you to connect it to a database somewhere else. A Helm7 Product already contains its services, database, auth, storage, domains and variables, so asking it a question gets the whole answer.

  1. Organisation
  2. Acme

Product

Environment production, preview, development

  • Serviceweb, worker or cron
  • DatabasePostgres 18
  • Authsign-in for your users
  • StorageS3-compatible buckets
  • Domainwith its certificate
  • Secretwrite-only variables

Environments hold copies of the same shape: production, a development environment, and on Pro, preview environments.

Deleting a Product removes everything under it, after a plan that lists what goes and a final backup of each database.

Costs roll up the same way, so the bill says what each application costs rather than what each vendor charged.

Each part, working.

Pick a part to see what it looks like inside a Product. Everything here can also be done over the API, the CLI or MCP.

Every operation in the API reference

Built from git. One call to roll back.

  1. Acme
  2. production
  3. web

Deployments of web, newest first

  • a1f3c2eAdd the pricing pageLive
  • 9bd01aaFix the login redirect
  • 77e4b19First deploy

Build log for a1f3c2e

  1. Cloned main
  2. Detected Next.js, npm, Node 22
  3. No Dockerfile in the repository, so one was generated
  4. npm ci
  5. next build
  6. Started the new container and waited for it to answer
  7. Moved traffic to the new container

What happens when you deploy.

Traffic moves only to a version that has answered. The version it replaces keeps serving until then.

  1. 1

    Build

    Clone the commit, detect the framework, and build an image. A repository with its own Dockerfile uses it; otherwise one is generated.

  2. 2

    Start

    Start the new version beside the old one and wait until it answers a health check. A cold framework boot gets 30 seconds of grace.

  3. 3

    Switch

    Move traffic to the new version only once it is healthy. A build that never answers never takes traffic.

  4. 4

    Drain

    Give the old version 30 seconds to finish the requests it already has, then stop it. The previous deployment is kept for rollback.

Four ways in, one surface.

The REST API, the MCP server and the SDK are generated from one catalogue of 66 operations, so none of them can offer something the others cannot. The CLI has a command for every operation: 65 commands in all.

DashboardFor people: your Products, what they are made of, and what they cost.Sign inREST API66 operations, with dry runs, idempotency keys and stable error codes.API referenceMCPFor agents. Tools a key cannot use are left out of its tool list.Agent accessCLIhelm7, with 65 commands, and --dry-run on the ones that change infrastructure.CLI reference

What it does not do.

Better to find these here than three weeks into a migration.

  • Helm7 runs in two EU regions today: Falkenstein (eu-central) and Helsinki (eu-north). There is no US region yet.
  • Services run in containers on a shared kernel. Dedicated nodes are a Scale-tier option, not the default.
  • Next.js is the best-supported framework today. Anything with a Dockerfile runs, but the detection, the build defaults and the documentation are thinner elsewhere.
  • There is no edge runtime, no GPU, and no Kubernetes to configure. Services run as containers.
  • Helm7 is new. There is no large community, no third-party plugin ecosystem, and no long public operating record to point at.

Questions about the platform.

  • What is the difference between PaaS and BaaS?

    A platform as a service runs your application code. A backend as a service hosts what the code depends on, such as the database, auth and file storage. Most teams use one of each; Helm7 is both, on one bill.

  • What can I run on Helm7?

    Next.js is the best-supported framework today. Anything with a Dockerfile runs, but the detection, the build defaults and the documentation are thinner elsewhere. Services can be web servers, background workers with no time limit, or cron jobs.

  • Where does it run?

    Falkenstein, Germany and Helsinki, Finland. There is no US region yet.

  • How isolated is my application?

    Services run in containers on a shared kernel. Dedicated nodes are a Scale-tier option, not the default.

  • What does it cost?

    From $12 a month for one application in production. Every price is a minimum spend that usage draws down against, and there is no per-seat charge.