One API, 200+ models
OpenAI-compatible and Anthropic-compatible gateways, one key, one balance. Same pricing as the app, nothing per operation.
Drop-in for clients that already speak OpenAI or Anthropic. Zero refactor to point at idapt.
Every provider behind one credential. Switch models per request, not per integration.
One quota across every provider. Per-model envelopes still enforced, but you do not track them yourself.
Cap a key by model, per-period spend, or rotation date. Revoke without restarting clients. A flat 5% fee applies only when you add credits.
How it works
Mint a key in Settings → API Keys, point your SDK at the gateway, and spend credits at the prices shown in the app.
import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://idapt.app/v1",
apiKey: "uk_...", // your idapt API key
});
const reply = await client.chat.completions.create({
// any model id from the in-app catalog
model: "...",
messages: [{ role: "user", content: "Hello" }],
});OpenAI compatible
https://idapt.app/v1
Any OpenAI SDK, LangChain, LiteLLM: works unchanged.
Anthropic compatible
https://idapt.app/anthropic/v1
Drop-in for the Anthropic SDK and Claude Code.
API pricing is the app pricing
Credits settle every model and paid tool at the prices shown in the app. The only fee is a flat 5% top-up fee when you add credits; nothing is charged per operation.
FAQ
Building a consumer product instead? See plans · Developer docs