Planned region → Canada
Not available yet. US East (New Jersey) is a region we are planning, not one you can deploy to. Hostim runs one live region today, in Falkenstein, Germany.
FastAPI on Hostim runs as a permanent container with model weights on a persistent volume, so nothing reloads between requests and streaming responses hold open as long as they need to. This runs in Germany today. A US East region in New Jersey — about 12 ms from Montreal — is planned. Join the waitlist if you would run inference there.
# docker-compose.yml
services:
api:
image: my-fastapi-app
command: uvicorn main:app --host 0.0.0.0 --port 8000
db:
image: postgres:16Montreal is one of the denser machine learning cities in the world, and a lot of the applied work coming out of it ends up as a FastAPI service wrapping a model. That workload has an awkward relationship with serverless: cold starts reload weights, timeouts truncate long generations, and per-request pricing penalises exactly the streaming behaviour you want. A long-lived container avoids all of it by being unremarkable — start the process, load from a mounted volume, stay up. Two limits are worth being straight about. Hostim is CPU-only today, which covers embeddings, classical ML and small models but not large-model serving, and GPU is roadmap rather than shipped. And the only live region is in Germany, roughly 95 ms from Montreal, which is a real cost when a client makes several sequential calls per user action. New Jersey would cut that to about 12 ms.
Expected latency. The expected site is Secaucus, New Jersey; the exact facility is not fixed and any northern New Jersey site gives similar figures. The figures below are typical round-trip times for that metro — they are estimates for the location, not measurements from Hostim hardware, because there is none there yet.
| City | Expected RTT to US East |
|---|---|
| Montreal | ~12 ms |
| Toronto | ~15 ms |
| Ottawa | ~16 ms |
| Halifax | ~25 ms |
| Winnipeg | ~40 ms |
| Calgary | ~55 ms |
| Vancouver | ~70 ms |
Privacy law. PIPEDA at the federal level, Quebec Law 25 (the strictest of the set, with transfer assessments and breach reporting), and Nova Scotia's PIIDPA, which still requires public-sector data to be stored and accessed only in Canada. British Columbia dropped its blanket residency rule in November 2021, so BC public bodies now decide foreign storage by their own policy rather than by statute.
Who enforces it. the Office of the Privacy Commissioner of Canada (OPC), and the Commission d'accès à l'information for Quebec.
Data residency — read this before you plan a migration. US East is in New Jersey, not Canada. If your contract or your provincial rules require Canadian storage, this region will not meet that requirement — tell us on the waitlist form and we will count it as demand for a separate Canadian region rather than pretending US East solves it.
Market context. Toronto and Montreal are 12–15 ms from New Jersey, so a US East region serves most Canadian users well. The exception is public-sector and health work: Nova Scotia requires storage inside Canada by statute, parts of the Quebec public sector do too, and many BC public bodies still require it by internal policy even though the statutory rule was removed in 2021. A New Jersey region does not satisfy any of those. Private-sector SaaS under PIPEDA generally can use a US region with the right contract terms.
What you are probably comparing us against: Heroku, Render, Railway, DigitalOcean, OVHcloud Canada, Vultr Toronto.
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
No, it is planned. The description on this page is of a region we are considering building, and the waitlist is the evidence we are using to decide.
Not in any region today. CPU inference with generous RAM works for embeddings, classical ML and smaller transformer models. GPU demand is worth recording on the waitlist form even though we cannot commit to it.
About 12 ms round-trip to northern New Jersey, against roughly 95 ms to Germany today. For a service that makes several chained calls per request, that difference compounds.
No, when they are on a persistent volume mounted at /models. The volume reattaches to the new container, so a deploy restarts the process without redownloading the model.
It can, depending on your institution and your funding terms — university and hospital agreements often impose residency conditions stricter than PIPEDA. Check before you migrate, and tell us if a Canadian region is the only thing that would work for you.
We are deciding how much capacity to build. Add your name and what you would run — that is the whole input. You can also deploy to our live German region today if that works for you.
Platform figures from the published price list, docs and our own benchmark.
€2.50/month
Hostim entry app plan
Plan sa-1-1: 1 vCPU, 1 GB RAM. Flat price, billed hourly, no usage meter.
€0
Managed database free tier
PostgreSQL 256 MB, MySQL 256 MB, Redis 128 MB, volume 1 GB. No cap on how many free instances you create.
Included on every plan
Database high availability
PostgreSQL and MySQL run as a primary plus hot standby with automatic failover, shared plans included.
2,708 TPS
Managed PostgreSQL write throughput
pgbench, 4 clients, 300 s, plan drp-50 (2 vCPU / 4 GB). AWS RDS db.t4g.medium scored 1,080 TPS on the same test.
€0/GB
Traffic charges
No ingress fees and no per-GB egress line on app plans.
1 today (Falkenstein, Germany)
Regions
Bare metal, EU only, with more regions planned. No AWS, GCP or Azure underneath.
“Managed PostgreSQL runs as a replicated cluster with automatic failover on every plan, shared and dedicated alike.”
— Hostim.dev docs, managed PostgreSQL