Rapids is DanubeData’s serverless container platform, built on Knative. Deploy an image, a Git repository or a ZIP archive, then manage resources, scaling, TLS and deployments through the console or CLI. It suits HTTP APIs, webhooks, web applications and internal services.

How does scale-to-zero work?

Set minimum replicas to 0 to let a service scale down after traffic stops and its scale-down delay has elapsed. The default delay is five minutes and can be configured. Cold-start time depends on the image, application startup and available capacity. Set minimum replicas to 1 or more to keep the service warm. Storage charges continue for any separately billed storage, even when compute scales to zero.

How does billing work?

Compute and requests are metered in vCPU-seconds, memory GiB-seconds and request count. Monthly allowances are deducted before charges on paid resource profiles. The Free resource profile has no compute or request charge while its monthly request count stays within the request allowance. Traffic, separately billed storage and any team pricing contract can affect the final bill.

Do I need a payment method for the free tier?

A payment method is required to deploy a Rapids container, including on the Free resource profile. The monthly allowances and rates shown on this page come from current public pricing.

Which deployment methods and languages are supported?

Use a Docker image, a Git repository from GitHub, GitLab or Bitbucket, or a ZIP upload. Source deployments can use Cloud Native Buildpacks or a Dockerfile. Buildpack language support depends on the selected builder; a Dockerfile gives you control over the runtime and system dependencies.

Can I use custom domains?

Yes. Each container has a public endpoint, and you can add up to 10 custom domains. Configure the required DNS records, then Rapids verifies the domain and manages its TLS certificate.

What are the resource and scaling limits?

Choose Free, Micro, Small, Medium or Large. The Free profile allows up to 3 replicas. Paid profiles have a platform maximum of 100 replicas, but profile and team quotas determine what can actually run. Concurrency can be configured up to 1,000 requests per instance.

Can an AI agent deploy and diagnose a service?

Yes. The danube CLI supports JSON output, image preflight, public endpoint probes and deployment diagnostics. Rapids apply supports idempotency keys. Operation status exposes an explicit terminal flag; operations wait can follow work to completion. Scope the token to the access your agent needs and review changes that affect your workloads.

Where do Rapids containers run?

Rapids containers run on DanubeData’s infrastructure in Falkenstein, Germany. Connections to external APIs, registries or destinations follow your application and deployment configuration.

Rapids.
Containers that scale with demand.

Deploy an image, connect a repository, or upload your source. Rapids handles builds, HTTPS and traffic-based scaling on EU infrastructure.

Create an account Choose your deploy method

EU hosted
TLS encrypted
German data centers

A payment method is required to deploy, including on the Free profile.

A deployment your agent can verify

Check that your image can be pulled before deploying it. Replace the example image with your own.

Preflight

danube rapids preflight --image ghcr.io/your-team/api:v1 --json

Apply the service and wait for the operation. Reuse the same idempotency key when retrying the same request.

Deploy

danube rapids apply --name api --image ghcr.io/your-team/api:v1 --wait --idempotency-key api-release-v1 --json

Probe an endpoint your application serves. A completed operation and a successful HTTP response answer different questions.

Verify

danube rapids probe api --path /healthz --cold --warm-requests 5 --json

Example commands. Authenticate and select your project before deploying.

What customers say

“As a developer and cloud enthusiast, I’ve tried a lot of infrastructure platforms — but DanubeData genuinely stands out as one of the most intuitive, modern, and reliable cloud services I’ve used. From signup to deployment, everything feels thoughtfully designed and developer-centric. Deploying a VPS, cache, or database instance takes under a minute — and that’s not an exaggeration! Their intuitive dashboard and instant provisioning make launching new services effortless, even if you’re just getting started with cloud infrastructure”

Keoma WrightCEO/Founder at OpenWeb.co.za

“I’m running my new project’s backend on DanubeData and the experience has been smooth from day one—deployment is fast, the dashboard feels genuinely developer-friendly, and everything has been stable with no issues so far. Any time I had a question or hit a small blocker, the support was excellent: quick to respond, clear in explanations, and focused on getting me unblocked rather than sending me in circles. If you want EU-based infrastructure without the usual DevOps friction, DanubeData delivers.”

Cristian TorzsaCEO at InnovaSoftware SRL

“ruby/Sinatra invoicing application on a Docker container. Danube data was an obvious choice given its incredibly user friendly dashboard and automatic deployment directly from my GitHub repository via webhook. I've used many other platforms (Google, AWS, Azure), but Danube is the clear winner in terms of ease to configure, clear and honest pricing not to mention the excellent response time. Everything works perfectly so I hardly ever have to pull up the dashboard since I moved to Danube Data.”

Bo PereiraParis-Brest srl

Three paths to your next deployment

Keep the workflow that fits your code. Every path produces a container revision you can inspect.

Docker image

Deploy a tagged image from a supported public or private registry. Set the port your application listens on and configure registry credentials when needed.

Explore Docker image

Git repository

Connect GitHub, GitLab or Bitbucket. Build with Cloud Native Buildpacks or a Dockerfile, choose a branch, and enable webhook deployments on push.

Explore Git repository

ZIP upload

Upload your source archive through the console or deploy a local directory with the CLI. Buildpacks or your Dockerfile turn the source into a container image.

Explore ZIP upload

From source to a managed service

Choose your source
Bring a container image, connect a Git repository, or upload source code. Configure the application port and environment variables.
Build and deploy
Rapids builds source deployments and creates a revision with a public HTTPS endpoint. Check its status, events and logs from the console or CLI.
Set your scaling policy
Choose a resource profile, minimum and maximum replicas, and a requests-per-second or concurrency target. Set minimum replicas to 0 to allow scale-to-zero.

The controls your service needs

Choose how each service runs, scales and receives traffic.

Traffic-based scaling
Scale using requests per second or concurrent requests. Replica counts remain subject to your resource profile and team quotas.
Scale-to-zero or stay warm
Minimum replicas of 0 allow idle services to stop. Keep at least one replica for workloads where a cold start would be disruptive.
Managed HTTPS
Rapids manages TLS certificates for its public endpoint and verified custom domains. Your application still owns its authentication and access rules.
Custom domains
Add up to 10 domains per container, configure the required DNS records, and track verification and certificate status in the console.
Deploy on a Git push
A webhook triggers builds for your selected branch. Inspect build logs and revision status before relying on a new deployment.
CLI diagnostics
Preflight an image, probe the public URL, or diagnose a failed deployment. JSON output gives an agent structured results to act on.

Work from your terminal or your agent

Preflight an image, apply a deployment, and check the public endpoint. Explore the Rapids and authentication commands in this playground.

Agentic · AI workflows

~/fleetAgentic · AI workflows

danube registry verify-push

Confirm a push is permitted before running it

SYNOPSIS danube registry verify-push <repository> --json

GET/v1/registry/verify-push?repository=acme%2Fapi

200OK

0ms260 B

application/json

// waiting for command…

SDK@danubedata/cli

AuthBearer $DANUBE_TOKEN

$ npm install -g @danubedata/cli · Explore the danube CLI → ·Free & open source · MIT

Structured responses

JSON output returns success, data, error and meta. Error codes and retryability help an agent choose its next action.

Explicit completion

Check the operation’s terminal flag or use operations wait. A wait timeout means the work may still be running.

Safe retries

An idempotency key on Rapids apply lets you retry the same request after losing the response.

Read the Rapids automation contract or use the machine-readable docs index.

Metered usage with monthly allowances

CPU, memory and requests use the same public rates across resource profiles. Your usage and any pricing contract determine the final bill.

Public usage rates

CPU: €0.054 per vCPU-hour
Memory: €0.009 per GiB-hour
Requests: €0.15 per million requests

Hour rates are shown for readability; compute is metered in seconds. Traffic and separately billed storage are additional.

See all pricing

Monthly compute and request allowances

2,000,000 requests
250,000 vCPU-seconds
500,000 GiB-seconds

Allowances are deducted before charges on paid profiles. The Free profile has no compute or request charge while it stays within the monthly request allowance.

Compare resource profiles

Free0.01 to 0.1 vCPU64 to 128 MBUp to 3 replicas
Micro0.1 to 1 vCPU256 to 512 MBLight APIs and bots
Small0.5 to 1 vCPU256 to 512 MBStandard APIs
Medium1 to 2 vCPU512 MB to 1 GBData processing
Large2 to 4 vCPU1 to 2 GBCompute-intensive work

Ranges show resource requests and limits per replica. Profile and team quotas cap scaling. Read the serverless documentation before selecting your configuration.

Rapids questions, answered

Scaling, deployment and billing details to check before you ship.

Need help with a workload? Read the documentation or talk to us.

Rapids is DanubeData’s serverless container platform, built on Knative. Deploy an image, a Git repository or a ZIP archive, then manage resources, scaling, TLS and deployments through the console or CLI. It suits HTTP APIs, webhooks, web applications and internal services.
Link to this answer
Set minimum replicas to 0 to let a service scale down after traffic stops and its scale-down delay has elapsed. The default delay is five minutes and can be configured. Cold-start time depends on the image, application startup and available capacity. Set minimum replicas to 1 or more to keep the service warm. Storage charges continue for any separately billed storage, even when compute scales to zero.
Link to this answer
Compute and requests are metered in vCPU-seconds, memory GiB-seconds and request count. Monthly allowances are deducted before charges on paid resource profiles. The Free resource profile has no compute or request charge while its monthly request count stays within the request allowance. Traffic, separately billed storage and any team pricing contract can affect the final bill.
Link to this answer
A payment method is required to deploy a Rapids container, including on the Free resource profile. The monthly allowances and rates shown on this page come from current public pricing.
Link to this answer
Use a Docker image, a Git repository from GitHub, GitLab or Bitbucket, or a ZIP upload. Source deployments can use Cloud Native Buildpacks or a Dockerfile. Buildpack language support depends on the selected builder; a Dockerfile gives you control over the runtime and system dependencies.
Link to this answer
Yes. Each container has a public endpoint, and you can add up to 10 custom domains. Configure the required DNS records, then Rapids verifies the domain and manages its TLS certificate.
Link to this answer
Choose Free, Micro, Small, Medium or Large. The Free profile allows up to 3 replicas. Paid profiles have a platform maximum of 100 replicas, but profile and team quotas determine what can actually run. Concurrency can be configured up to 1,000 requests per instance.
Link to this answer
Yes. The danube CLI supports JSON output, image preflight, public endpoint probes and deployment diagnostics. Rapids apply supports idempotency keys. Operation status exposes an explicit terminal flag; operations wait can follow work to completion. Scope the token to the access your agent needs and review changes that affect your workloads.
Link to this answer
Rapids containers run on DanubeData’s infrastructure in Falkenstein, Germany. Connections to external APIs, registries or destinations follow your application and deployment configuration.
Link to this answer

Explore more solutions

Give your next service a home

Bring your image or source code, choose a scaling policy, and manage the service with the console or your agent.

Create an account Review pricing

Start free

Compute

Storage

Managed apps

Caching

Databases

Messaging

Monitoring

DanubeData CLI

Automation & Agents

Tools

Developer docs

Learn

Support

Rapids.Containers that scale with demand.

A deployment your agent can verify

Docker image

Git repository

ZIP upload

Choose your source

Build and deploy

Set your scaling policy

Traffic-based scaling

Scale-to-zero or stay warm

Managed HTTPS

Custom domains

Deploy on a Git push

CLI diagnostics