AI features without per-token pricing.
The calls your app makes for its users: summarise this thread, answer from our help centre, tag this upload, draft that email. Ship the feature without pricing it per token.
Pre-launch. Accounts are open; the API and checkout open soon.
How it runs.
01
Call it from your backend.
Any OpenAI or Anthropic SDK works: change the base URL and the key.
02
Pick a model per feature.
Quick models for tags and summaries, a stronger one for answers that have to be right.
03
Size lanes to your peak.
Count the calls in flight at your busiest moment. That’s your lanes.
Why lanes fit it.
A cost for your spreadsheet.
A flat monthly price instead of a bill that moves with every user’s habits.
Heavy users are fine.
The customer who summarises everything costs you the same as the one who never clicks the button.
Try another model in one line.
Every model is on every paid plan, so switching is a change to the model name.
Works with what you use.
AI SDK
typescript
import { createOpenAICompatible } from "@ai-sdk/openai-compatible";import { generateText } from "ai"; const voidstone = createOpenAICompatible({ name: "voidstone", baseURL: "https://api.voidstone.net/v1", apiKey: process.env.VOIDSTONE_API_KEY,}); const { text } = await generateText({ model: voidstone("kimi-k3"), prompt: "Write a migration for this schema change.",});