WHERE WE DEPLOY
We optimise every layer of inference. And deploy it wherever you say.
The same layer, the same team and the same optimisations run in three different places depending on what your case needs. Starting in one and moving to another is not a migration: it is a deployment decision.
The route
The same stack, three destinations.
On our nodes
Our own infrastructure in Spain. You manage nothing: not hardware, not stack, not scaling.
api.nextbit256.com · Running
Regioneu-south · EU
Uptime 90 d99,9 %
Node local time--:--:--
Serverless · Dedicated
In your cloud
We deploy and operate our stack inside your AWS, Azure or GCP account. You leverage your spend commitment and your network policies.
Runs inYour account on
RegionThe one you choose
eu-west-1us-east-1me-central-1
Final availability depends on the hyperscaler.
Managed Dedicated Inference
On your premises
Real on-premise: physical hardware in your datacenter, zero data egress. Operated end to end by us.
Sovereign deployment
Decision table
What changes between them.
| On our nodes | Dedicated / Capacity block | On your premises | |
|---|---|---|---|
| Isolation | Multi-tenant with logical isolation | GPU dedicated in hardware | Fully physical |
| Jurisdiction | Spain (Nextbit) | Spain (Nextbit) | Yours |
| Who operates it | Nextbit, everything | Nextbit, everything | Nextbit operates, you host |
| Pricing model | Per million tokens | Fixed monthly fee | Project + fee |
The same API in all three. Anyone who starts on serverless and grows into dedicated or on-premise does not rewrite a single line of integration.