One endpoint · every model · pay per token

Every AI model,
billed by the token

One OpenAI-compatible endpoint for hundreds of models. No subscription — you pay purely for what you use, per million tokens, per model. Bring your apps, keep your pricing.

# Point any OpenAI SDK at Idexal
client = OpenAI(
  base_url="https://api.idexal.com/v1",
  api_key="sk-idexal-...",
)
client.chat.completions.create(
  model="claude-opus-4.8", messages=[...]
)
Three steps

From key to call in minutes

1

Get an API key

Sign up, top up credit, and generate a key. Your key, your usage, your control.

2

Point your SDK at us

Swap the base URL to api.idexal.com/v1. Any OpenAI-compatible client just works.

3

Pay for what you use

Every call is metered by tokens. Watch spend live in your dashboard; no surprises, no plans.

Usage-based pricing

Per million tokens. Per model.

Transparent, metered pricing — you're charged for the tokens you send and receive, nothing more.

Model
Input / 1M
Output / 1M
claude-opus-4.8
$15.00
$75.00
claude-sonnet-4.6
$3.00
$15.00
gpt-5.5
$5.00
$20.00
gemini-3.5
$2.00
$10.00
+ 750 more models
see dashboard
see dashboard

Indicative pricing — final per-model rates are set at launch. No subscription, no minimums.

Built for production

Everything you'd expect — and more

Hundreds of models

Claude, GPT, Gemini, and 750+ more — one endpoint, always up to date.

Automatic fallback

If a model errors or is busy, requests fail over to a backup — your app keeps running.

Live usage dashboard

Track spend, tokens, and per-model usage in real time. Set limits and alerts.

Keys & team management

Issue scoped keys per project or teammate; rotate and revoke any time.

OpenAI-compatible

Drop-in for the OpenAI SDK and tools you already use. No rewrite.

Powers Idexal Code

The same gateway behind our own coding agent — production-proven.

Start building on every model

One key. One endpoint. Pay only for the tokens you use. Get started the moment we launch.