10 Arma Reforger Hosting Providers, Compared Honestly

Search for Arma Reforger hosting and you will find a dozen ranked lists. Almost all of them are published by a hosting company, or paid for by one — affiliate commission is the business model of most "best host" round-ups, so the order often reflects who pays most per signup rather than who runs the better server. Almost all of them put the publisher or the best-paying partner first, and almost all of them rank on price per slot.

This post is no exception on the first count: we are a hosting company, and we are one of the ten providers ranked below. On the second count it is — nobody paid to be here, and nobody paid us. So let us deal with that first, then deal with the bigger problem, which is that price per slot is close to meaningless for this particular game.

How this was put together#

The disclosure. Terabit hosts Arma Reforger servers. This post exists partly because we would like you to consider us. You should read it with that in mind, which is exactly why the method is here at the top rather than buried at the bottom.

Concretely: none of the outbound links below are affiliate links, and we earn nothing if you read this and pick one of the other nine. That is also why several of them are recommended over us further down.

What was checked. Every price below was read from the provider's own page in August 2026. Where a provider does not publish a readable price, it says so rather than repeating a number from an affiliate site — several of those sites disagree with each other, which is a decent sign none of them checked either.

The criteria, in the order we weighted them:

1. Does the pricing model match how Reforger actually consumes resources?
2. Is protection against attacks included, or an add-on?
3. How much Arma-specific tooling is there, versus a generic panel with Reforger bolted on?
4. Location coverage — specifically a good route to where the players are, not the size of the list.
5. Whether you can find out the price without starting a checkout.

Those criteria put us first, which is exactly what you would expect a hosting company's own ranking to do — so here is the honest shape of it. We win on 2 and 3. We lose on price, to the provider immediately below us. Weight cost above resilience, as plenty of people reasonably should, and the top two swap. The table has enough detail for you to re-sort it yourself, which is rather the point of publishing it.

The problem with every other comparison#

Here is what nobody selling you a slot wants to spend long on.

These providers are not selling the same unit. GTX Gaming, Host Havoc, LOW.MS and PingPerfect sell player slots. BisectHosting sells gigabytes of RAM. GPORTAL sells a fixed package. Qonzer and Terabit sell CPU threads, memory and storage that you set independently. Ranking those by headline price is arithmetic on four different things.

And for Reforger, the slot is the wrong unit. The server's load is dominated by how many AI units it is simulating and how far it is streaming entities to players — not by how many humans are connected. A 32-player Conflict session with AI uncapped will work a server harder than 64 players with the AI limit set. We wrote the sizing logic up in full, and the short version is that operating.aiLimit and serverMaxViewDistance decide your hardware needs far more than your player cap does.

So when a host sells slots, the main thing it can sell you more of is a bigger player cap — usually with resources attached that you cannot decline separately.

The table#

Prices as listed on each provider's own site, August 2026. Three list in GBP, so an approximate USD figure is shown alongside — comparing a pound price against a dollar one by eye is how people talk themselves into a difference that is not there. Conversions move with the exchange rate. Check current pricing before you buy; this is the fastest-ageing page we publish.

Three of the ten will not show you a price without entering their order flow. That is not a scandal, but it does mean any "cheapest host" list including them is guessing.

The ten, and who each is actually for#

1. Terabit#

Ours, and we have put it first, so weigh that accordingly. Same resource-based model as Qonzer at a higher entry price — the sections below are what the difference buys. We host the majority of the top NA and EU Reforger communities, by both player count and server count, having entered the game in November 2025.

Price#

$13.65/mo at entry, with CPU, RAM and storage set independently and no slot cap. This is the criterion we lose on, and it is the one below us that wins it.

One pool, several servers#

Because you are buying resources rather than a slot cap, you can carve them up. Run the public Conflict server and a training server out of the same pool, move CPU and RAM between them when the player counts shift, and spin a third up from whatever is spare instead of buying a second plan. It stays one invoice and one renewal date however many you end up running, and configs, kits and ranks can be kept in step across all of them.

The same applies when the group moves on for a while. A server can be switched to any other game in our catalogue from the panel — it keeps its IP, its storage and its place on your bill, and the old files stay on disk, so coming back to Reforger for the next update picks up where you left off rather than starting again.

Hardware#

Every server runs on a Ryzen 9 7950X or 9950X. There is nothing older in the fleet, which is a different claim from listing a fast CPU on the premium tier and quietly running something else underneath the cheap one. It matters more in this game than most: Reforger's simulation tick is largely serial, so clock speed sets your server frame rate, and a per-slot plan rarely tells you what you are actually running on.

DDoS mitigation#

We specialize in this area, with an in-house fully-owned network capable of 13+ Tbps of stateful filtering, and 500+ Tbps of volumetric capacity through three upstream partners.

Included at every tier rather than sold as one, with a filtering profile built and tuned for Arma Reforger's own traffic rather than a generic game profile. The largest attack it has absorbed against a Reforger server is 12 Tbps — that figure is from this game, on our own network, not borrowed from an unrelated target, which is worth checking whenever any host quotes you a number.

The model is layered, and deliberately not single-sourced. Volumetric capacity of 500+ Tbps comes through three upstream partners — Cloudflare, GSL and CDN77 — rather than one, so trouble with capacity or routing at any single provider does not take the whole footprint with it. Underneath that sits 13+ Tbps of stateful filtering we run in-house, on hardware we own, which is what makes a new attack pattern a deploy rather than a ticket in another company's queue.

We have written up how that filtering was built, including the one attack vector that did get past it — a low-bandwidth flood imitating real game clients, in March 2026, closed in under two hours. Nothing has got through since. We would rather you read the version with the failure in it than take a marketing line on trust.

Mod tooling#

This goes well past a Workshop installer, and for most communities it is the part that changes the day-to-day more than anything above.

• Dependencies are ordered for you, rather than left to a load order you discover is wrong when a mod silently fails.
• Whole modlists save as collections, optionally carrying the scenario with them, so switching between a milsim night and a public rotation is one action.
• Version locking means a Workshop update cannot quietly change your server on its next restart.
• A collision guard catches the duplicate mod name that otherwise stops every mod on a server loading while the config still looks perfectly correct.
• When a server will not start, it works out which mod is responsible and offers you the fix, instead of leaving you to bisect a sixty-mod list by hand on a Friday night.

If you run mods at all, it is the difference between administering a server and fighting it.

Insights#

Server FPS, frame time, entity load and player ping, charted over time next to the usual CPU and memory. The part that matters for this game is that AI, vehicles and projectiles are separate lines, so "the server felt bad last night" resolves into a setting you can actually change rather than a shrug — which is the same argument the sizing article makes, with your own numbers behind it.

Crashes are recorded and grouped, so something that keeps happening reads as a pattern rather than a series of unrelated bad evenings — each group carries what kind of crash it was, the engine build it happened on and how much memory was in play, which is usually enough to separate a misbehaving mod from a server that has simply run out of room. And Bohemia's backend status sits alongside your server, because when their services are down nothing about your server is wrong and no amount of restarting it will help.

Community#

Who is actually playing, and whether they come back. Retention, session length, average ping, where players connect from, and a kill feed with friendly fire separated out — the numbers you need when you are deciding whether a rule change worked.

It also flags the connections worth a second look: VPN, proxy, Tor and datacentre joins, and accounts that cluster together in the way ban evasion tends to. You choose which of those get turned away and how strict the threshold is — the ones you block never make it into your server. Everything else here is evidence for moderators rather than a moderation panel, which is the way round we think it should be: the calls stay yours.

Locations#

One caveat on that locations column, ours included: those are each provider's advertised network locations, which is not the same as where you can put this game tonight. Arma Reforger currently deploys with us in Montreal, New York, Dallas, North Kansas City, Amsterdam, London and Frankfurt, with more of the network being brought online for it. Ask any provider on this list the same question before you buy, because the advertised number rarely answers it.

The honest limit#

We are not the cheapest. The provider below undercuts us on the same billing model, by roughly three dollars a month at entry once the currencies are lined up.

2. Qonzer#

The cheapest entry price we could verify on a resource-based model, and the closest direct comparison to us on this list — same unit of sale, different tradeoffs. Worth reading against the section above rather than on its own.

Price#

£7.95/mo for a CPU thread, 4 GB and 40 GB of NVMe, with no slot limit — and double the storage we give you at entry. They are cheaper, though by less than the currencies suggest: £7.95 is roughly $10.75, against our $13.65. Call it three dollars a month between two similar entry tiers.

Hardware#

Stated on their own site: their hardware rests on a "primary partnership with Limestone Networks" rather than being their own. That is a perfectly normal way to build a hosting company and plenty of good ones do it. It is worth knowing only because it changes who you are talking to when something at the machine level has to change.

DDoS mitigation#

Third-party as well, per their own pages — NSFocus appliances, with GCORE on the paid tiers — and the volumetric layer rests on that single upstream rather than a blend of several. Protection is tiered, with the stronger levels charged monthly on top of the server.

Two consequences worth weighing rather than dismissing. A novel attack pattern is a conversation with a vendor rather than a change they can make themselves, which is the distinction criterion 2 was about. And depending on one upstream means one set of capacity and one set of routes to depend on — fine on almost every day, and precisely the case a blend exists to cover.

Locations#

24 advertised, second only to ours in this table, and broader than any of the slot-based providers. Treat that the way you should treat everyone's here, ours included. What decides whether a server feels good is a clean route to your players, and every provider on this list covers North America and Western Europe, where the overwhelming majority of Reforger communities actually are. A longer list mostly adds places like Australia and South America — genuinely useful if that is where your players sit, and largely decorative if it is not. Bandwidth in those regions is expensive, which is why coverage there is patchy across the whole market rather than a quirk of any one host.

The honest limit#

If price is your binding constraint and you are not expecting to be attacked, they are the cheaper way to buy the same shape of server, and we would rather say so.

3. GTX Gaming#

The lowest European starting price at £9.99/mo for 10 slots, with 19 locations and a Workshop mod installer in the panel. Sensible if you are EU-based, want a small server tonight, and per-slot pricing suits how you think about it.

4. XGamingServer#

$10.50 to $42.00 across five tiers on Ryzen 9 7950X hardware, which is the right shape of CPU for a simulation bound by single-thread speed. The standout is documentation — they have written more publicly about Reforger server administration than anyone else on this list, which is worth something when you are stuck at 11pm. Only five locations, so check one is near your players.

5. LOW.MS#

$10.95/mo for 10 slots, working out at $1.09 a slot, across 12 locations with crossplay and Workshop support stated. Straightforward, and the per-slot maths is the easiest to reason about here.

6. GPORTAL#

£17.19 per 30 days for 20 slots as a fixed package. Prepaid 30-day billing rather than a rolling subscription suits people who would rather not have a recurring card charge. Primarily a European footprint.

7. Host Havoc#

From $20.00/mo, starting at 40 slots and going to 128, across 14 locations with offsite backups and their own hardware. The entry tier is larger than most groups need, so this makes sense once you are actually filling 40 slots.

8. BisectHosting#

Prices per GB of RAM behind a configurator, with 21 locations and one of the largest game catalogues in hosting. Swapping a server from one game to another is not unusual any more — we do it, and so do others here — but the sheer breadth of what you can swap to is genuinely theirs. If your group works through a long tail of smaller titles between Reforger patches, that is the reason to pick them, and it is a real one.

9. PingPerfect#

Per-slot from 10 slots, price only inside the configurator, but the location list is the most unusual here — Johannesburg, São Paulo, Santiago. If your community is somewhere the big providers ignore, this is the list to check first.

10. Nitrado#

Long-established and widely used for console game hosting, on per-slot prepaid packages. We could not read pricing or slot ranges from their site to verify anything, so this placement reflects what we could confirm rather than a judgement on the service. Worth a look if your group is console-first.

Pick someone else if…#

• You want the lowest price on a resource model — Qonzer, though the gap at entry is about three dollars a month rather than the gulf the price tags imply.
• You are in Europe and want the cheapest slot-based start — GTX Gaming.
• Your group works through a long tail of other games — BisectHosting, for catalogue breadth. Keeping one server and changing the game on it is common now, this list included; the number of games you can change it to is what actually differs.
• Your community is in Australia, South America or Africa — PingPerfect reaches the most unusual places on this list, and Qonzer is broad too. Coverage outside NA and Western Europe is thin across the whole market, so this narrows the field fast.
• You are console-first — Nitrado is the one built around that audience.

Where we would argue for ourselves: you run mods and would rather spend the evening playing than ordering a JSON array, you want to know the CPU under your server without reading a spec sheet twice, or you are big enough that being attacked is a question of when. The first of those covers most communities, which is why we did not put ourselves first on the DDoS argument alone.

What to ask before you buy#

Whoever you end up with, these are the questions that separate providers faster than any price table. The answers below are ours; the point is to ask everyone the same thing.

1. 1

   Is DDoS protection included, or is it a paid tier?

   What to listen for: A straight answer about what the free level actually stops. Several hosts include a basic level and charge monthly for anything that holds up under a real attack — which is fine, as long as you knew before you were being attacked.

   Terabit.io: Included on every server, at every tier. The model is layered: 500+ Tbps of volumetric capacity through upstream partners, plus 13+ Tbps of stateful filtering we run in-house. The largest attack we have absorbed against a Reforger server specifically is 12 Tbps.

2. 2

   Who writes a new filter when a novel attack gets through?

   What to listen for: Whether the people who can fix it work at the host or at an upstream vendor. If the filtering belongs to someone else, a new attack pattern is a support ticket in another company's queue.

   Terabit.io: We do, on our own network and our own hardware. That is the reason the Reforger filtering exists at all — it came out of defending communities attacked frequently enough to expose what generic mitigation misses, and owning the stack is what makes a fix a deploy instead of an escalation.

3. 3

   Can I change CPU, RAM and storage independently?

   What to listen for: Whether upgrading one resource forces you to buy the others. On a per-slot plan, more CPU usually means buying a bigger player cap you did not want.

   Terabit.io: Yes, each is set separately, and the configurator tells you when more CPU threads will not help rather than selling them to you.

4. 4

   What happens to my modlist when a mod updates on the Workshop?

   What to listen for: Whether you can lock a mod to a specific build. Without it, any restart can quietly pull a new version and stop your players joining a server that looks perfectly healthy.

   Terabit.io: Lock any mod to its current build, or leave it on latest. Collections save whole modlists — optionally with the scenario — so switching setups is one action.

5. 5

   How many player slots do I actually need?

   What to listen for: A host that asks about your AI settings and scenario before answering. One that only asks how many players you have is selling you a number, not sizing a server.

   Terabit.io: Fewer than most people buy. Cap your AI first and read the frame time — we wrote the whole sizing argument up rather than putting it in a sales conversation.

The honest summary#

If you take one thing from this: do not shop on price per slot for this game. Work out what your scenario and AI settings demand, then find the provider whose pricing model lets you buy that without paying for a player cap you will never fill.

Several providers on this list will do that well. We think we do it best for anyone running mods seriously, on hardware picked for the way this game actually behaves — and we have tried to be specific about where that argument stops applying, which is on price.

New to running one? Start here — the setup path, config reference and troubleshooting are all free to read, whoever you buy from.

Sources#

Pricing and feature data read from each provider's own public pages in August 2026, in line with how we source the comparison tables elsewhere on the site. Providers change prices without notice — confirm before you buy.

• Qonzer · GTX Gaming · XGamingServer
• LOW.MS · GPORTAL · Host Havoc
• BisectHosting · PingPerfect · Nitrado