UTIO

AI API

AI API: transcription, voice, text and OCR with a single key

One key instead of six providers. Transcription, voice, text, image, OCR and embeddings with one AI key; UTIO's CRM and channels with a separate channel key; all on the same services our products run on in production. For developers and agencies who want to charge for what they build.

Open my panel account
$ curl -X POST "https://studio.utio.io/api/v1/transcribe" \
    -H "Authorization: Bearer $UTIO_API_KEY" \
    -F "file=@call.mp3"

{ "text": "Hi, I'd like an appointment on Thursday…" }

In short

UTIO's AI API gives you transcription, text to speech, text, image, OCR and embeddings with one AI key, and UTIO's CRM and channels with a separate channel key, on the same services our own products run on.

  • Two ways to pay: a prepaid wallet (€50 minimum top-up) or a panel account at €29/mo + usage. No commitment.
  • Keys with the client on record and a spending cap, revocable instantly, with a limit of 120 requests per minute per key.
  • Channel keys require an Idempotency-Key: a retry never duplicates a message.
  • Reselling it under your brand? That's white-label Studio.

Last reviewed:

What the API does: nine capabilities, one API

No endless catalogue: we publish what is in production. Each capability is a route under the same base and the same authentication. Also OpenAI-compatible (swap base URL and key) and available as an MCP server for Claude Code, Cursor and other MCP clients that connect with a key.

Endpoint What it does
/transcribe Transcription. Calls and audio to text, with timestamps per sentence and, on request, per word.
/tts Text to speech. Natural audio from text, in several languages and voices.
/text Text. Replies, summaries and classification, with the model tier of your choice.
/image Image. Images from a description.
/ocr OCR. Scanned invoices, delivery notes and forms to structured text.
/embeddings Embeddings. Vectors to index and search your own content.
/contacts CRM. Contacts, timeline, pipelines, opportunities and tasks in UTIO's CRM, readable and writable from your application.
/whatsapp/send requires Idempotency-Key Channels. Send over WhatsApp and email with dedicated channel keys.
/sales/documents Sales. Quotes and invoices ready for VeriFactu (Spain).

Real governance, not a checkbox

Tokens with the client on record

Every token is created in the panel with the client on record: you know which key spends, how much and for whom. Instantly revocable.

Spending cap and request limit

Every key carries its own spending cap, which you set and change in the panel, plus a limit of 120 requests per minute per key.

Idempotency-Key on channels

Channel keys require the Idempotency-Key header: a network retry never duplicates a message, a contact or an invoice.

What it costs: two ways to pay

You pay for what you use. No commitment on either.

Prepaid wallet

€50 minimum top-up

No monthly fee. Top up and consume. For trying the API and for per-project billing.

Panel account

€29/mo + usage

Per-client keys, spending reports and support. For agencies and products in production.

Per-unit rates to four decimal places by model tier (basic, standard or advanced), published in the panel.

How you call it

Authentication via the Authorization header with your panel key. Minimal bodies: just enough to get a response.

Transcribe a call (curl)
curl -X POST "https://studio.utio.io/api/v1/transcribe" \
  -H "Authorization: Bearer $UTIO_API_KEY" \
  -F "file=@llamada.mp3"
Text to speech (JavaScript)
const res = await fetch("https://studio.utio.io/api/v1/tts", {
  method: "POST",
  headers: {
    Authorization: `Bearer ${process.env.UTIO_API_KEY}`,
    "Content-Type": "application/json",
  },
  body: JSON.stringify({ text: "Tu cita queda confirmada.", voice: "Kore" }),
});
const audio = await res.arrayBuffer();
Send over a channel (curl + Idempotency-Key)
curl -X POST "https://studio.utio.io/api/v1/whatsapp/send" \
  -H "Authorization: Bearer $UTIO_CHANNEL_KEY" \
  -H "Idempotency-Key: 6f1c9b2e-el-mismo-en-cada-reintento" \
  -H "Content-Type: application/json" \
  -d '{"phone": "+34600000000", "message": "Tu cita queda confirmada."}'

In the panel: your keys, your usage, a quick start and the integrator guide for the OpenAI-compatible API.

Would rather have the agent ready-made than build it? See the voice agent, the WhatsApp agent and the email agent, ready to use.

Reselling it under your own brand?

If you are an agency and want the panel, the rate card and the wallet under your domain and logo, that is white-label Studio.

See white-label Studio

AI API FAQ

How much does UTIO's AI API cost?

Two options. A prepaid wallet with no monthly fee and a €50 minimum top-up. Or a panel account at €29/mo plus usage, with per-client keys and spending reports. Per-unit rates go to four decimal places by model tier and are published in the panel. No commitment on either.

What can I do with the API?

Transcribe calls and audio, turn text into speech, generate and classify text, create images, turn scanned documents into text with OCR and build embeddings to search your own content. You can also read and write UTIO's CRM and send over WhatsApp and email.

How do I control what each client spends?

Every key is created in the panel with the client on record and its own spending cap, with a limit of 120 requests per minute. You know which key spends, how much and for whom, and you can revoke it instantly.

What happens if my app retries a send?

Channel keys require the Idempotency-Key header. If you replay the request with the same key, the message isn't duplicated.

Where are the docs?

In the panel: your keys, your usage, a quick start and the integrator guide for the OpenAI-compatible API. This page has minimal examples in curl and JavaScript.

Can I resell it under my own brand?

Yes. If you're an agency and want the panel, the price list and the wallet on your own domain and logo, that's UTIO Studio as a white label.

Tell us what you are going to build

Volume, the capabilities you need and whether you resell to clients: and we'll get back to you with the model tier that fits.

  • 🔒 Secure data
  • ⚡ Reply within 24h
  • ✓ No commitment

No commitment · Reply within 24 working hours · We only use your details to reply to you.

UTIO is the AI agent line of Universal Telecom, an operator registered with the CNMC.