Arc Cloud Is Live: Managed Arc, No Infrastructure Required
Six months ago I wrote the hello world post for Basekick Labs. I talked about where Arc came from — the computer I bought in 2022 for side projects, the failed Wavestream experiment, Historian, and eventually Arc.
At the end of that post I said: "Let's make history."
Today, we're taking the next step.
Arc Cloud is live.
How We Got Here
When we shipped Arc's first public commit in October 2025, the goal was simple: prove there was hunger for a time-series database that separated storage from compute and didn't make you sign an NDA to read the pricing page.
The response surprised us. 358 GitHub stars in the first month. POCs with healthcare companies. Engineers running Arc in production before we even called it stable.
But there was a consistent piece of friction in every conversation.
"I love Arc. How do I run it?"
And the answer was: you pull a Docker image and you're up in five minutes. Arc is genuinely easy to self-host — one binary, one container, no config files required.
But some teams don't want to manage infrastructure at all. Not because it's hard — because it's not their job. They want to send data and query it, not think about storage backends, upgrades, or uptime.
We kept hearing it. From IoT teams who wanted to start collecting sensor data the same afternoon. From observability engineers who wanted to evaluate Arc before committing to a self-hosted deployment. From founders who had enough infrastructure to manage already.
So we built Arc Cloud.
What Arc Cloud Is
Arc Cloud is managed Arc. Same binary, same DuckDB engine, same Parquet storage, same SQL. We run the infrastructure. You run your queries.
You sign up, pick a plan, and 60 seconds later you have an Arc instance with:
- A dedicated endpoint (
your-id.arc.us-east.basekick.net) - HTTPS out of the box
- Automatic hourly compaction
- A console to run SQL queries from the browser
- Storage monitoring with overage alerts
No infrastructure to manage. No upgrades to schedule. No on-call rotation for your database.
Just Arc.
Infrastructure
Arc Cloud runs on bare metal — dedicated Velia servers, not shared virtual machines. Your instance gets real CPU and real NVMe, which is a big part of why Arc's performance numbers hold up in production.
We're launching in Miami (US East) today. Frankfurt is coming soon for European customers — data stays in the EU, which matters for GDPR compliance and latency. If you're in Europe and want to be notified when Frankfurt goes live, email hello@basekick.net and we'll reach out the day it opens.
What's Different from Self-Hosted
Nothing about Arc itself changed. The open source version is still there, still AGPL, still the same binary.
What changed is who runs it.
When you self-host, you own the infrastructure. That's powerful — no one between you and your data, full control, Arc running exactly how you want it.
Arc Cloud is for the teams who'd rather spend that energy elsewhere. Uptime, storage, upgrades, monitoring — that's our problem, not yours.
Your data is still in Parquet files. You can export it anytime. No lock-in — that's non-negotiable for us.
Pricing
There's a free trial so you can actually try it before you decide anything.
Paid plans start at $50/month. The price scales with CPU, memory, and storage — not with how much data you ingest or how many queries you run. We think per-query pricing is a tax on curiosity, and we're not doing that.
See all plans at basekick.net/cloud
Why Now
I'll be honest — we could have launched Arc Cloud three months ago.
We didn't because I wanted to make sure the core was right first. The 26.04.1 release we shipped last week included a full security audit, 9 critical fixes, a native Arrow query path that made queries 59-157% faster, and automatic deduplication in compaction. That's the foundation Arc Cloud runs on.
I didn't want to put "managed Arc" in front of people and then have the underlying product bite them. The open source releases had to be solid before the cloud offering made sense.
They are now.
The Part That Keeps Me Up at Night (In a Good Way)
We have real workloads running on Arc Cloud already. Weather sensor data pushing every 10 seconds. Kubernetes monitoring. Application logs. People using the Grafana datasource against their cloud instances.
It works. The compaction runs on schedule. The queries are fast. The dashboard shows storage usage in real time.
But we're still early. We're a small team, and this is our first managed offering. If something breaks, we want to know immediately. Reply to any email you get from us, drop into Discord, open a GitHub issue — whatever works for you. We read everything.
Get Started
Start a free trial. No credit card required to try. Takes about 60 seconds to get your first instance running.
If you're already self-hosting Arc, nothing changes for you — and if you ever want to move to managed, your data is in Parquet and you can bring it with you.
Resources:
Questions? hello@basekick.net — goes straight to me.
— Ignacio
Founder, Basekick Labs
Ready to handle billion-record workloads?
Deploy Arc in minutes. Own your data in Parquet. Use for analytics, observability, AI, IoT, or data warehousing.