AI Coding Plans

One key. Every model. A monthly token bank.

Point any OpenAI-compatible client at a single endpoint and reach the whole catalog. Tokens reset every month, and every plan includes a dedicated VPS with full root.

Five plans, one endpoint

Each plan pairs a monthly token allowance with a VPS sized for the same workload.

125M tokens / month

Go

$30.00/ month

  • 4 GB RAM · 2 vCPU
  • All models, one key
  • Up to 1M-token context
  • OpenAI-compatible /v1
Order Go
250M tokens / month

Plus

$60.00/ month

  • 16 GB RAM · 8 vCPU
  • All models, one key
  • Up to 1M-token context
  • OpenAI-compatible /v1
Order Plus
500M tokens / month

Pro

$125.00/ month

  • 32 GB RAM · 16 vCPU
  • All models, one key
  • Up to 1M-token context
  • OpenAI-compatible /v1
Order Pro
1B tokens / month

Ultra

$250.00/ month

  • 64 GB RAM · 32 vCPU
  • All models, one key
  • Up to 1M-token context
  • OpenAI-compatible /v1
Order Ultra
2B tokens / month

Max

$500.00/ month

  • 128 GB RAM · 64 vCPU
  • All models, one key
  • Up to 1M-token context
  • OpenAI-compatible /v1
Order Max

Token allowances reset monthly. Unused tokens do not roll over.

How the endpoint works

OpenAI-compatible

Change the base URL and the key in your existing client. No SDK rewrite, no proxy shim to maintain.

POST /v1/chat/completions

Context up to 1M tokens

Long repositories and large documents fit in a single request on models that support the window.

Per model

Pipe-width throughput

Published TPS figures describe the ceiling of the route, not a sustained guarantee for every call.

Varies by model

Model routing without the guesswork

Heavier models draw down the token bank faster than light ones, so you can keep routine work on a fast model and spend the heavier routes on the problems that need them.

Requests may be served by self-hosted capacity or by an upstream provider configured for that model. Either way you authenticate once, against the same key.

Try it against your own workload

The catalog and the live status of every route are public. Check the model you need before you buy.

Check live model status