EU hosting → Germany
German FastAPI teams running ML APIs or async backends want low latency, persistent volumes for model weights, and DSGVO compliance. Hostim runs FastAPI in Falkenstein in long-lived containers — no cold starts, no model reload between requests.
# docker-compose.yml
services:
api:
image: my-fastapi-app
command: uvicorn main:app --host 0.0.0.0 --port 8000
db:
image: postgres:16FastAPI is the default modern Python web framework for ML inference and async APIs. The deployment shape is Uvicorn + Gunicorn (or pure Uvicorn) in a Docker container. ML APIs typically need a multi-GB volume for model weights — Hostim provides persistent volumes that survive deploys. German enterprise customers often ask whether the model weights leave the EU; on Hostim they sit on a Falkenstein disk and never replicate elsewhere.
Latency. Typical RTT from Berlin to our Falkenstein region is around 8 ms. Germany has the strictest interpretation of GDPR in the EU and is the largest EU market for paid hosting. Falkenstein is in Saxony, Germany — Hostim runs entirely on bare metal in this region.
Regulator. BfDI (federal data protection commissioner) and BSI (federal cyber security agency).
Local law. BDSG (Bundesdatenschutzgesetz), the German federal data protection act that complements GDPR. Hostim is operated by HOSTIM.DEV UG, a German company, with all data in Falkenstein, Germany — there is no transfer outside the EU for application data.
Local alternatives you may have considered: IONOS Cloud, Strato, OVH-DE, Hetzner Cloud (raw VPS).
FastAPI hosting means running a Uvicorn or Gunicorn-Uvicorn worker process and exposing it over HTTPS. The framework is async by default, so the host has to support long-lived connections — websockets, server-sent events, streaming responses.
Hostim runs your FastAPI Docker image as a normal container. Long-lived connections work. Managed PostgreSQL is attached at runtime. If you serve an ML model, mount a persistent volume for the weights so they do not redownload on every deploy.
Cold starts from serverless platforms are not a fit for ML inference workloads. Hostim runs a permanent container, so model weights stay in memory across requests.
DATABASE_URL, OPENAI_API_KEY, MODEL_PATH, LOG_LEVEL
Nein, wenn du sie auf einem Persistent Volume speicherst. Mounte ein Volume an /models, lade die Gewichte einmal beim Build oder beim ersten Start, und sie bleiben auch nach jedem Redeploy.
Yes. Hostim runs your FastAPI container as a normal long-lived process behind HTTPS — server-sent events and websockets work without extra config.
Aktuell nein. CPU-Inferenz mit ausreichend RAM funktioniert für viele Use Cases (kleinere Embeddings, klassisches ML). GPU steht auf der Roadmap.
Enable managed PostgreSQL in the project; Hostim injects DATABASE_URL into the FastAPI container. Async drivers (asyncpg) work out of the box.
Spin up an app in minutes. Managed database on the free tier, custom domain included.