Token API
Shared API access, served from your chosen UK or EU region. No commitment and no minimum.
- Budgets and attribution
- Per-team and per-key hard caps
- OpenAI-compatible API
Pricing.
Metered, reserved or private capacity. Ask for a quote based on your workload.
Dedicated Endpoints
Your model on GPUs no one else uses. Run on demand or reserve capacity for a committed term.
- Bring your own weights
- On-demand scale to zero
- Reserved capacity options
Private Nodes
Physically dedicated nodes, with serving, monitoring and patching managed by us.
- Dedicated serving stack and cache
- Private networking
- Customer-held encryption keys
Measured on your traffic.
Tell us your model, region and expected workload. We’ll help scope the deployment and quote.