Catalog
Breadth is the signal. We still split on-demand vs spot and we still want GPU + region on the row.
Provider review / verified 2026-09-05
FluidstackGPU review
Fluidstack sits closer to a broker/managed marketplace than to a single-datacenter neocloud. Breadth is the pitch. The work is knowing whose machine you landed on.
Models listed4
From /GPU/hr$0.44
Regions in sample1
Business modelManaged GPU cloud
Our reading
Who this provider is for
I open Fluidstack when I want coverage, models and regions, without opening twelve host UIs. I do not confuse that with CoreWeave. You are still shopping capacity that may live in more than one kind of facility. Read the GPU, the region, the interconnect, then the hour. If the hour is the only thing you read, you are back on Vast.ai with extra steps.
Rent it if People who want a wide GPU menu through one account and will still inspect the row.
Skip it if Buyers who need a single operator SLA and a named fabric. Go to a neocloud.
How we compare hoursStrengths
- Wide model coverage relative to single-site clouds
- One workflow when you are tired of N marketplaces
- Can surface regional options you would miss on a US-only shop
- On-demand listings we can put next to RunPod and Vast.ai
Limits
- Operator identity is fuzzier than Lambda or CoreWeave, know that going in
- Quality of a given hour depends on the underlying site
- Not the place I send InfiniBand cluster romantics by default
- Support path can feel like a broker, because parts of it are
Same standard, every cloud
What we measured on Fluidstack
Four checks we apply to every provider. This is not a score, and it is not a paid ranking.
Price signal
Often aggressive vs hyperscalers. Compare to Vast.ai and RunPod on the same GPU before celebrating.
Product shape
Managed access to distributed GPU inventory. Dashboard first, fabric second.
Still verify
Region, interconnect, and who operates the node. The brand on the tab is not the whole story.
- Model
- Managed GPU cloud
- Headquarters
- London, UK
- Setup
- Dashboard / managed GPU access
- Coverage
- Multi-site inventory
- On-demand
- Yes, as listed
- Founded
- 2017
- Best known for
- Broad GPU access
- Compare with
- Vast.ai, RunPod, Hyperstack
Catalog listings
Fluidstack catalog sample
Published on-demand hours first. Spot and waitlists are labeled. Confirm the live rate before you provision.
Brokers are allowed. Confusion is not.
There is nothing immoral about aggregating GPUs. There is something expensive about thinking you bought a single SLA. I treat Fluidstack as a panel. Each listing is a bet. The dashboard is convenience. Convenience is not interconnect.
Where it wins
When I need an awkward SKU or a region the usual suspects do not show. When I want one invoice. When Vast.ai’s host UI is too much chaos and RunPod does not have the card.
When I need a 256-GPU IB island, I stop and call a neocloud.
How I actually rent it
- 01
Filter GPU + region
Do not shop ‘a GPU.’ Shop this model in this place.
- 02
Read interconnect and stock
PCIe vs NVLink vs IB. In stock vs waitlist.
- 03
Keep a backup cloud
RunPod or Lambda if the aggregated node misbehaves. Two accounts is cheaper than a dead epoch.
FAQ
Fluidstack questions
Is Fluidstack like Vast.ai?+
Spiritually closer to a managed marketplace than to a single cloud. Less DIY host-vetting than Vast.ai, less operator-clarity than Lambda.
Is Fluidstack good for H100?+
When they list it on-demand, yes, as a row to compare. Confirm the site behind the row.
Who should use Fluidstack?+
People who want breadth and one dashboard, and who will still read the listing.
Fluidstack vs Hyperstack?+
Both chase ‘easy GPU cloud’ from Europe. Compare the live SKU, not the accent.
Sources
Where these numbers come from
Prices and SKUs are from the GPUBeacon catalog sample. Ranking rules are on the methodology page. Outbound links may be affiliate. They never change the hour we display.
GPUs
Keep comparing
Other providers to compare
Same fields, different product. Compare the GPU hour before you lock a brand.
