> For the full page index, fetch https://echonova.ai/llms.txt

# What is EchoNova

Arabic-first speech and conversational AI, in the browser and over an API.

EchoNova is an Arabic-first speech and conversational AI platform. Use it from the
portal in your browser, or call the same capabilities directly from your own
product with the Developer API.

## What you can do in the portal

- **Voices.** Browse a library of Fusha (Modern Standard Arabic) and dialect
  voices, and preview them before you use them anywhere else.
- **Text to Speech.** Turn written text into natural speech. An "Add
  diacritics" (tashkeel) button lets you mark or correct the diacritics that
  drive Arabic pronunciation before you generate audio.
- **Speech to Text.** Upload audio and get a transcript back. Short files
  transcribe synchronously; longer files run as an async job you can poll or
  get notified about.
- **Sound Cleaning.** A voice isolator that strips background noise from a
  recording so the speech underneath is clear.
- **Assets.** Every voice, generated clip and transcript you create is kept
  here until you delete it, so you can find and reuse it later.

## Agents

An [agent](/docs/agents/what-is-an-agent) pairs a persona with a knowledge base
(documents you upload plus anything crawled from a website you point it at) and
a set of tools it can call: webhooks, OpenAPI actions, database lookups, and
system tools. You connect an agent to the channels your customers already use:
web chat and web calls, WhatsApp, Messenger, Instagram, Slack, SMS and voice
calls over Twilio, SIP trunking, Genesys, Zendesk, Zoho Desk and Intercom.
Agents can also reach into a Google Calendar to check or book time.

## Workspaces and billing

Work happens inside a workspace, which holds your voices, agents, channels
and billing. Usage is metered
with credits, and the portal shows what each action costs as you go.

## The Developer API

Text to Speech and Speech to Text are also an API. Create an API key in the
portal, then call the `/v1/text-to-speech*`, `/v1/speech-to-text*` and
`/v1/sound-cleaning*` endpoints (fifteen in total) from your own backend. The
[Developer API overview](/docs/developers/overview) lists every endpoint and
the pages that explain them. Webhooks notify you when an asynchronous job
finishes, and the request log shows every call your key made.

## FAQ

### Do I need to write code to use EchoNova?

No. Everything in the portal works in the browser. The API is for teams
that want speech inside their own product.

### Which languages does it speak?

Fusha (Modern Standard Arabic), Arabic dialects and English, one per voice.
Transcription is in Arabic.

### Where do I start?

Pick a voice under **Voices**, generate a sentence under **Text to Speech**,
then read [Get an API key](/docs/developers/get-an-api-key) when you are
ready to call it from code.

## Related

- [Base URL](/docs/developers/base-url): the host your
  API key talks to
- [Get an API key](/docs/developers/get-an-api-key): your first
  authenticated request
- [What is an agent](/docs/agents/what-is-an-agent): persona, knowledge and
  tools
- [Developer API overview](/docs/developers/overview): all fifteen endpoints
- [Glossary](/docs/help/glossary): every word the portal uses with a meaning
  of its own