xs
trying things out
€4/ mo
€0.0055 / hour
take xs- vcpu
- 1
- memory
- 1 gb
- nvme
- 10 gb
- port
- 50 mbit/s
- traffic
- 1.5 tb
one bot, one cron, nothing heavy
serversthatdon'tgetslowerwhensomeoneelsegetsbusy.
no synergy, no hyperscale, no 99.99% you can't verify. just memory that stays yours, a person who replies, and prices you can check against the provider we buy from.
looz console
4 machines · parismachinecpustatus
- web-01203.0.113.17 · plan m12%running
- db-main203.0.113.18 · plan l34%running
- worker-02203.0.113.21 · plan s6%running
- staging203.0.113.24 · plan s0%stopped
example: node load, synthetic data28%
an illustration, not a measurement of our machines
a mock-up of the panel, not live data. the real one shows your machines and what each costs per hour.
plans and prices(01)
five sizes. no hidden line items.
you pay by the hour and stop paying the moment you delete the machine. a public ipv4 and a monthly traffic quota are in every plan — the quota is on the card below; over it, €1 per extra terabyte, and we say so here rather than in an invoice.
s
side projects
€7/ mo
€0.0096 / hour
take s- vcpu
- 1
- memory
- 2 gb
- nvme
- 20 gb
- port
- 1 gbit/s
- traffic
- 2 tb
bots, cron jobs, small apis
m
most people
€12/ mo
€0.0164 / hour
take m- vcpu
- 2
- memory
- 4 gb
- nvme
- 40 gb
- port
- 10 gbit/s
- traffic
- 3 tb
websites, databases, staging
l
small teams
€22/ mo
€0.0301 / hour
take l- vcpu
- 2
- memory
- 8 gb
- nvme
- 80 gb
- port
- 10 gbit/s
- traffic
- 5 tb
production apps, ci runners
xl
growing products
€42/ mo
€0.0575 / hour
take xl- vcpu
- 4
- memory
- 16 gb
- nvme
- 160 gb
- port
- 10 gbit/s
- traffic
- 9 tb
heavy workloads, several services
prices exclude vat · billed by the hour, delete any time · pay in usdt (trc-20)
what's included(02)
everything you need is in the base price
always included
- dedicated ipv4 address
- a traffic quota that grows with the plan — 1.5 tb on xs, 9 tb on xl
- nvme storage
- hourly billing, delete any time
- full root access, any linux distro
nothing else to sell you
there are no paid extras. we had a list of them here — extra addresses, monitoring, managed maintenance — and took it down because none of it existed behind the price. when one does, it comes back with a button, not with a line on this page.
why looz(03)
three things you can verify, not three adjectives
no overselling
your memory is yours. we don't resell the same gigabyte three times and hope nobody notices at 8pm. cpu is shared the way it is everywhere, but nothing can push your machine into swap.
on our own hardware, a public load graph — today the machines are rented
12h
a human answers
within twelve working hours, from the person who runs this — not a ticket queue and not a chatbot with three canned replies. miss the window and the month is free.
telegram @looz_support, and a ticket in the panel that starts the clock
snapshots on demand
you snapshot the disk yourself, from the machine page, in about a minute, and it costs nothing. what we do not have is an automatic backup — so we do not bill you for one. a copy you have been charged for and never seen is worse than none.
a button on the machine page, not a ticket
support(04)
small enough that a message reaches an actual engineer
big hosts replaced people with chatbots and call it 24/7. we can't match that theatre, so here's what actually happens instead.
you write
telegram @looz_support, or a ticket from the panel. no priority tier, no bot asking you to describe the issue in a dropdown.
a person reads it
the same person who set up the hypervisor and holds root on it. nothing gets escalated to a second line, because there isn't one.
it gets fixed
within twelve working hours — 09:00 to 19:00 CET, monday to friday, the same window the service agreement names. if we miss it, that month costs you nothing: no form to fill in, we just don't bill it.
paris · amsterdam · riga
we grow one machine at a time.
the honest bit(05)
you're probably tired of hosting pages promising enterprise-grade infrastructure, 99.99% uptime nobody measures, and 24/7 support that turns out to be a chatbot with three canned answers.
so here it is plainly. we are small. we are small enough that you get a person, not a queue — and we would rather say what we do not have yet than sell you a promise we cannot keep. there is no global edge network, no ai-powered anything, and no wall of customer logos — because we're new and putting fake ones there would tell you everything about how we'd handle your data.
what you get instead: a price you can check against the provider we rent from, a bill counted in hours rather than months, and a snapshot you can take without asking permission. if that's worth six euros more a month than the cheapest host you found, we should talk.
if it isn't, that's a completely reasonable decision and we won't retarget you across the internet for six weeks.
faq(06)
straight answers
why are you more expensive than the cheapest hosts?
because we allocate memory one-to-one instead of packing more customers onto the same box. if price is your only criterion, a host charging €2 will always beat us — and your server will be slow at 8pm.
where are the servers?
the machines run in datacentres we rent capacity in, and we are the only party you deal with. europe — paris, riga, amsterdam — plus hong kong and three us locations. every plan runs on nvme drives. each plan includes a monthly traffic quota — from 1.5 tb on the smallest to 9 tb on the largest, and it is written on the plan card. going over is billed at €1 per extra terabyte by the provider whose capacity we rent, and we pass that on unchanged. we cannot yet show your traffic in the panel, so we are not going to promise a warning we have no way to send.
what happens if a disk fails?
disk redundancy belongs to the provider whose cloud the machine runs on, and we are not going to invent a layout we cannot inspect. what is ours to say: there is no automatic backup here. take a snapshot before anything risky — it is a button on the machine page and it costs nothing — and keep your own copy of whatever you cannot lose.
how fast do i get my server?
minutes. the panel orders it through the provider's api as soon as your balance covers a day of running, and the machine boots while you are still on the page.
can i pay with crypto?
yes — usdt (trc-20) is how you pay here, and for now the only way: no cards yet. you top up a balance and we bill by the hour. we do ask for a verified contact — anonymous signups are what attract spam and get whole networks blacklisted.
what isn't allowed?
spam, phishing, port scanning, ddos participation, and the rest of the list in the acceptable use policy. the machines sit in several countries, so the rule does not hang on any one country's law. outgoing mail ports are closed by default and opened on request once you have paid for a while.
can i move to a bigger plan later?
yes, and it takes minutes. you keep the same ip address, so nothing on your side needs reconfiguring.
do you offer refunds?
there is nothing to refund, because nothing is paid in advance: billing is hourly, and deleting the machine stops the meter that minute. if it does not work as described in the first week, we return what it cost you, without a conversation.
let's talk(07)
talk to a human
no forms, no card on file. tell us the size you want and what you're running — you'll have the server today.