Skip to content
Pricing

Free for builders. Paid for organizations.

Everything an individual needs is free, and publishing earns you more. Organizations pay for privacy, control and scale. Compute is never on our bill.

Free

$0forever

 

For every individual: build, evaluate, train and publish on your own stack.

Start free

Includes

  • Unlimited public environments and benchmarks
  • 10 private projects
  • 2 runners on your own machines
  • 50,000 traces a month
  • Datasets from traces: SFT, DPO, distillation
  • Evaluate and train with any model and trainer
  • Publish to the hub, post and claim bounties
  • Creator rewards: higher limits for what you publish

Business

For teams

$2,000per organization / month

billed monthly

For companies and labs working together in private.

Start 14-day trial

Everything in Free, for your whole team, plus

  • Unlimited members, roles and permissions
  • Unlimited private projects and runners
  • 1,000,000 traces a month
  • A private hub for your organization
  • Private leaderboards and private bounties
  • SSO with SAML
  • Audit log
  • Priority support

Enterprise

Custom

 

For organizations with security, compliance and scale requirements.

Talk to us

Everything in Business, plus

  • Runners inside your VPC
  • SCIM provisioning
  • Unlimited traces
  • Custom integrations and hubs
  • Uptime SLA and a dedicated engineer
  • Invoicing, security review and procurement
Creator rewards

Publish more, get more, for free.

People who upload are how the hub grows. Every environment or benchmark you publish that passes the audit raises your limits on the Free plan.

1Each audited public environment or benchmark+2 private projects, +1 runner, +25,000 traces a month
25 audited public listingsBusiness-level limits on your personal account
3Used by other teamsFeatured in the hub, with your name on every leaderboard it powers
On every plan

You never pay us for compute.

Models, GPUs and sandboxes are billed by your own providers at their prices. Plans pay for privacy, scale and control, never for compute.

RunPodLambdaAWSGoogle CloudModalE2BOpenAIAnthropicTogetherOpenRouter
Compare plans

Every limit, side by side.

Compute is never on the bill. Plans differ in privacy, scale and control.

Plan

Free

$0

Business

$2,000

Enterprise

Custom

Build

Public environments and benchmarks
Unlimited
Unlimited
Unlimited
Private projects
10 + rewards
Unlimited
Unlimited
Environments built by your coding agent
Traces a month
50k + rewards
1M
Unlimited
Datasets from traces (SFT, DPO, distillation)

Run

Runners on your own machines
2 + rewards
Unlimited
Unlimited
Any model, compute, sandbox and trainer
Runners inside your VPC
Compute markup
None
None
None

Share

Publish to the hub
Creator rewards
Post and claim bounties
Private leaderboards and bounties
Private organization hub

Admin and support

Members
Just you
Unlimited
Unlimited
Roles, permissions and audit log
SSO with SAML
SCIM provisioning
Support
Community
Priority
Dedicated
Your accounts

One bill per provider, all of them yours.

Connections live with your runner. Nenyax never holds a card for compute, and model keys never leave your machine.

  • Providers' prices, no markupGPUs, model calls and sandboxes are invoiced by the providers themselves.
  • Switch any timeSwap a provider by changing a connection; the environment never notices.
  • No credits to buyThere is no Nenyax compute balance to top up or run out of.

Your billswho bills you

RunPodGPUsyour account
OpenAIModel callsyour account
ModalSandboxesyour account
AWSStorageyour account
Nenyax compute$0
Questions

Good to know.

No. Models, GPUs and sandboxes are billed by your own providers at their prices. Nenyax never resells compute or model calls, on any plan.

Start free, on what you have.

A model endpoint and a machine are enough. Upgrade when you need privacy or scale.

Start free Talk to us