Native speed, no warm-up
Compiled Rust with no garbage collector means flat, predictable latency from the first request to the billionth.
Ferrum builds memory-safe, blazing-fast infrastructure in Rust: an edge runtime, an observability pipeline and a secrets vault that scale from a side project to a global fleet.
Why Ferrum
Performance and safety are not features you bolt on later. They are the foundation every product is built on.
Compiled Rust with no garbage collector means flat, predictable latency from the first request to the billionth.
The borrow checker removes whole classes of vulnerabilities before code ever reaches production.
Deploy to every region at once. Traffic is routed to the nearest healthy node automatically.
Structured logs, metrics and distributed traces are on by default and speak OpenTelemetry out of the box.
Small crates with clear contracts. Adopt one product or the whole platform, and swap parts as you grow.
A single tiny binary on a hobby server today, an autoscaled fleet tomorrow. Same code, same config.
Products
Start with the piece that solves today's problem. Everything shares the same control plane and the same ergonomics.
Run Rust and WebAssembly next to your users.
Ingest millions of events per second without blinking.
Secrets and keys that never leave safe memory.
Technology
One language, one toolchain and one type system across the whole stack. This very page is rendered by Leptos on an Axum server.
One language end to end keeps the team fast and the surface area small.
Server-rendered HTML for instant paint, hydrated in the browser for interactivity.
Async, modular HTTP services that handle enormous concurrency on modest hardware.
Everything is typed, validated at the edge and observable in production.
About us
We are engineers who got tired of firefighting and decided to build the tools that make it unnecessary.
We started Ferrum after one too many 3 a.m. pages caused by problems a compiler could have caught. Our founders spent a decade running large-scale systems and kept seeing the same failures: memory bugs, unpredictable latency, and tooling that told you what broke long after your customers noticed.
So we built the infrastructure we wished we had: written in Rust, observable from day one, and small enough for a single engineer to understand. Today we are a distributed team that ships in the open, writes thorough post-mortems and treats reliability as a product feature.
If the type system can prevent a bug, it should.
We design for the engineer reading the dashboard at night.
Our core crates are open source, and so are our roadmaps.
Pricing
Start free. Upgrade when your traffic does. No surprise overages, no long-term contracts.
For side projects and early experiments.
Free
Free forever, no card needed
Start for freeFor startups shipping to real customers.
$39/ month
Billed annually
Choose GrowthFor organisations with strict requirements.
Custom
Tailored to your needs
Talk to salesCustomers
Sample testimonials: replace these with quotes from your real customers.
We moved our ingestion tier over a weekend. P99 latency dropped by two thirds and our on-call rotation finally got quiet.
The compiler catches what code review used to. Shipping on Friday afternoon is no longer a scary sentence.
Observability that works the day you deploy. We cancelled two other tools in the first month.
FAQ
Can't find what you're looking for? Send us a message below and a human will reply.
No. Our products expose standard HTTP, OpenTelemetry and WebAssembly interfaces, so you can run workloads written in many languages. Rust gives you the best experience, and our guides will get you productive quickly.
Starter is free forever within its limits and needs no credit card. If you exceed them we pause new traffic rather than surprising you with a bill.
Yes. The core runtime and observability pipeline are open source. Managed plans add global routing, hosted storage and support.
In the region you choose. Enterprise plans can pin data to specific jurisdictions, and everything is encrypted in transit and at rest.
Most teams migrate incrementally: route a small percentage of traffic to Ferrum, compare, and cut over when you are ready. Our team will help plan it at no charge on Growth and above.
Cancel any time from your dashboard. You can export all of your configuration and data, and monthly plans are never charged for the following period.
Contact
Tell us a little about your project. We'll set up a call, a trial, or just answer your questions.