Pricing designed for
real AI builders

Transparent, flexible, and fair. Whether you’re compiling one wheel a month or hundreds a day, WheelForge fits your workflow.

All plans support pay-as-you-go builds. Cache hits are always free.

Free
£0 / month

Start for free. Pay only when you need GPU time.

  • Unlimited cache hits
  • 1 automated GPU build / mo
  • CLI + environment detection
  • Access to supported libraries list
  • Basic build logs
  • Pay-as-you-go builds enabled

Automated builds run at lowest queue priority on the Free Tier.

Pay-as-you-go GPU builds apply to all plans. You only pay for GPU time when monthly limits are exceeded.

Start for Free
Pay-As-You-Go
£0.20+ / build

Only pay for the GPU time you actually use.

  • A40 short build: £0.20–£0.30
  • L40S build: £0.40–£0.70
  • H100 full build: £1.00–£1.40
  • Cache hit = free
  • Available on every tier
Enabled by default
MOST POPULAR
Early Access
£12 / month

Be the first to use automated GPU builds with on-demand cloud GPU runners.

Status: Waitlist Open
  • 10 automated GPU builds / mo
  • Faster queue priority
  • Support for launch libraries
  • Environment snapshot + metadata
  • Vote-access on new library support
  • Early Access Discord
  • “Founding Developer” badge

Pay-as-you-go GPU builds apply to all plans. You only pay for GPU time when monthly limits are exceeded.

Join Early Access

Future Plans

Developer
£29 / month

For builders who need frequent GPU compilation.

Status: Planned (post-EA)
  • 60 GPU builds / mo
  • Faster queue priority
  • Warmed base images
  • Limited global cache access
  • Environment inspector (planned)

Pay-as-you-go GPU builds apply to all plans. You only pay for GPU time when monthly limits are exceeded.

Team
£99 / month

Build reproducible environments as a group.

Status: Planned
  • 300 GPU builds / mo
  • Shared team cache
  • Build retention policies
  • Notifications (Slack/Webhooks)
  • Team roles & permissions

Pay-as-you-go GPU builds apply to all plans. You only pay for GPU time when monthly limits are exceeded.

Enterprise
Custom

Full binary infrastructure for serious AI teams.

Status: Planned
  • Dedicated GPU runners
  • Custom CUDA configurations
  • On-prem or VPC deployments
  • Private binary registry
  • SBOM & provenance
  • SSO/SAML & SLAs

Pay-as-you-go GPU builds apply to all plans. You only pay for GPU time when monthly limits are exceeded.

Contact

Why this model works

Free Tier

Grows the cache by encouraging users to download existing wheels.

Pay-As-You-Go

Removes subscription friction for one-off builds.

Early Access

Gets automated builds into developers' hands early.

Enterprise

Shows ambition for full binary infrastructure scale.

Ready to start building?

Join the waitlist for early access to the GPU build engine.