Skip to content
State of GPU Compute 2026

Partner program

Put dedicated GPUs inside your offer.

Resell, embed or supply 8-GPU servers through Nexpod. Your customers get whole machines in EU and US regions. You skip the hardware, the on-call and the billing system.
Apply to partnerTalk to sales
partners.nexpod.ai/accounts
Sample data

Sub-accounts

4

GPUs live

104

Month to date

€125,856

  • Customer A4× B300 · EU-AMS-1€79,104
  • Customer B2× H200 · EU-FRA-1€20,352
  • Customer C6× B200 · EU-AMS-1€23,040
  • Customer D1× H100 · US-NYC-1€3,360
One invoice to you · EUR
in every server you sell
8GPUs
in every server you sell
for all your sub-accounts, in €
1invoice
for all your sub-accounts, in €
EU-AMS-1 · EU-FRA-1 · US-NYC-1
3regions
EU-AMS-1 · EU-FRA-1 · US-NYC-1

Partners building on Nexpod

Placeholder
  • Partner nameReseller
  • Partner nameIntegration
  • Partner nameCapacity
  • Partner nameHub publisher
  • Partner nameSystems integrator
  • Partner nameResearch

Why partner

Four reasons teams call us.

Most partners arrive with one of these on their desk. Each one maps to a track below.
  1. 01

    You resell compute to your own customers

    Put whole 8-GPU servers in your catalogue without buying one. You keep the relationship; we run the racks and the on-call.

  2. 02

    Your product runs on GPUs underneath

    Create servers and endpoints per customer from inside your product through the API, billed to you by the second.

  3. 03

    You own racks, power or GPUs

    Bring capacity into a Nexpod region. We burn it in, run the software layer and keep it rented.

  4. 04

    Your clients need EU residency

    Offer GPU capacity in EU regions, contracted and invoiced by an EU entity in euros, with data that stays in region.

Partner types

Six tracks. Pick the one you live in.

Each track has its own terms, tooling and contact on our side. Most partners start in one and grow into a second.
NODEYour hardware · 8× B200
8/8 online
0
B200
1
B200
2
B200
3
B200
4
B200
5
B200
6
B200
7
B200
Capacity

Data centers and GPU owners

Bring racks or servers into a Nexpod region. We burn them in, list them and keep them rented.

Statement · Octserver·h
  • Customer AB300 · 1,648
  • Customer BH200 · 636
  • Customer CB200 · 576
  • Customer DH100 · 140
  • Customer EH200 · 88
  • Customer FB300 · 64
Invoice to you · EUR6 accounts · on-demand rates€131,744

Your brand

on invoices

Billing

per second

Resellers & MSPs

Sell Nexpod as your own

Sub-accounts per customer, wholesale terms, one invoice to you.

await nexpod.servers.create({
  family: "B300",
  region: "eu-ams-1",
  tenant: customer.id,
})
Integrations

Build on the API

Provision servers per tenant from inside your product.

your-lab/model-70b

Nexpod Hub · template

verified
8× B200HTTP + streamEU-AMS-1
listed inHub · Inference
Model publishers

Ship models to Nexpod Hub

Verified templates that deploy on 8-GPU servers in one step.

Client workspaceshared with you

  1. 01

    Scope

    Workload and sizing

  2. 02

    Pilot

    2 servers, 2 weeks

  3. 03

    Production

    Reserved terms

Pilot capacity
2× H200 · EU-FRA-1
People
client · you · engineer
2-week pilot · on-demand
€21,504
Consultancies & SIs

Deliver GPU projects for clients

A shared workspace per client, pilot capacity and a partner engineer on call.

GPU quota · spring termorganiser view
  • Course A24 / 32 GPUs
    spend cap412 / 600 GPU·h
  • Lab B16 / 24 GPUs
    spend cap310 / 400 GPU·h
  • Hack C8 / 16 GPUs
    spend cap108 / 120 GPU·h
Research & education

Courses, labs and hackathons

Quotas per class or team, with spend caps the organiser controls.

By invitation

Structured agreements for larger partners.

Multi-year terms, named engineers and joint planning, for organisations that commit capacity or customers at scale.

  • Resellers and VARs

    Wholesale terms, your brand on the invoice

  • Managed service providers

    Operate Nexpod capacity for your clients

  • Venture funds and accelerators

    Credits and terms across a portfolio

  • Systems integrators

    Joint delivery on multi-node builds

  • Public sector and research

    EU residency, procurement-ready contracts

Apply

Tell us how you want to work.

Pick the tracks that fit and add a few details. A partner manager reads every application and replies within three business days.

01 · Pick your tracks1 picked

02 · Your details

How would you like to work together?

Used only to review your application. See the privacy policy. Demo form

  1. 03Review≤ 3 business daysWe check fit and confirm the track with you.
  2. 04Intro callWeek 1Thirty minutes with a partner manager and an engineer.
  3. 05Sandbox and termsWeek 2A partner account, credits to test and a draft agreement.

FAQ

What partners ask first.

The questions that come up before anyone signs.
Which track fits us?
Start from what you want your customers to get. Reselling and managed services sit in the reseller track; building on the API is an integration; bringing hardware is capacity. The intro call settles it in fifteen minutes.
How do partners earn?
Resellers buy at wholesale and set their own price. Referral and integration partners earn a share of the usage they bring in. Capacity partners are paid per rented server-hour. Exact terms come with the agreement.
Is there a fee or a minimum commitment?
Applying is free and there is no entry fee. Some tracks, such as capacity and structured reseller agreements, come with technical or volume requirements we agree on together before you sign.
Can our customers' workloads stay in the EU?
Yes. Pick EU-AMS-1 or EU-FRA-1 and data, backups and support access stay in that region. Contracts and invoices come from our EU entity, in euros.