NEW

GPU instances are live. Use code LOCAL10 for 10% off your first month.

Deploy now
about us

Operators,
not middlemen

Locally is built by people who were customers first: proxy operators, self-hosters and AI tinkerers who got tired of metered clouds and mystery hardware.

We started the way most of our customers did: with one open model, one underpowered machine, and one API bill that made no sense. The fix was obvious in hindsight. Stop renting tokens. Rent the machine.

So we racked hardware where the peering is good, wrote a provisioner that hands you root in minutes, and priced everything as one flat number a month. No meters, no throttling, no "fair use" asterisks. The whole company fits in one sentence: good hardware, honest specs, and we get out of the way.

Today we run 12 datacenters across three continents, operated by a small remote team between the Caribbean and Europe. Small is deliberate: every support reply comes from someone who can actually SSH into the problem.

0

customers and counting

0

servers online right now

0%

uptime, last 90 days

0

datacenters, 3 continents

0 min

average first support reply

$0

charged per token, forever

The longer version of the why lives in our manifesto. The short version: the most important technology of this decade should be something you can own.