Your agent also creates audio, voice, music and video (and this is what it costs)

What your agent can create

Besides texts, reports and admin tasks, your agent can generate real multimedia content:

• Spoken voice — turns any text into audio with a natural voice
• Sound effects — rain, applause, a door, whatever you describe
• Music — original pieces of whatever length you ask for
• Transcription — turns any audio into text, distinguishing who is speaking
• Voice cloning — with permission from the voice’s owner
• Video — clips generated from a sentence, or your photos turned into video

There’s nothing to set up: you ask for it in your usual chat, like any other task. The audio reaches you directly — on Telegram, as a voice note.

Spoken voice: turn any text into audio

Ask your agent to read a text to you, to turn an article into audio so you can listen to it on the go, or to prepare the voiceover for a video. It sounds natural — not the robotic voice of a few years ago.

There are different voice qualities: the standard one is fast and sounds great; if you need top quality (for example for material you’re going to publish), just ask and it will use the higher-quality mode, which costs a little more.

Sound effects and music

Need the sound of rain on a roof for a video? A short jingle for your podcast? Describe it to your agent and it generates it:

• Sound effects: from half a second to 30 seconds. The better you describe it, the better it comes out.
• Music: from 3 seconds up to 10 minutes. It’s charged per second, so ask for the length you really need.

Transcribing audio: from voice to text

Forward your agent that 4-minute audio you don’t feel like listening to, or the recording of a meeting, and it hands it back to you as text. It works in more than 90 languages and can tell who says what when several people are speaking.

Cloning a voice (with permission)

With one minute of clean audio, your agent can clone a voice and then use it to read texts. It’s useful for giving your own voice to your content without recording yourself every time.

Important rules, no fine print:

• Only with the explicit permission of the voice’s owner. Cloning someone else’s voice without their consent is not allowed.
• When you clone a voice you’ll receive a legal notice (your agent will pass it on to you and it also arrives by email to the account holder).
• Each agent keeps one cloned voice: if you clone another, it replaces the previous one.
• Only the Professional plan keeps the voice protected and persistent. On the other plans the cloned voice is a trial one and may be released if another agent needs the slot.

Changing the voice of an already-recorded audio

It can also take an existing audio and regenerate it with another voice — for example, you record a quick note on your phone and convert it into a more polished voice to publish it.

Video: from a sentence to a clip, or animating your photos

Two ways to create video:

• From a sentence: "a drone flies over a vineyard at sunset" → generated clip. There are several video models with different prices and qualities.
• Animating an image: you give it a photo and it turns it into a video with natural movement. There’s a budget option and a higher-quality option.

Keep in mind that video takes 1–3 minutes to generate — your agent will deliver it when it’s ready. The generated clips are short (around 8 seconds), designed for social media, websites and presentations.

How much each thing costs

Everything you generate comes out of your balance — there’s no separate fee and no "free" generations. Approximate prices in dollars:

To put it in context: adding a voice to this entire article would cost less than $1; a 3-minute song, about $0.90; and video is the most expensive — that’s why your agent will ask you for confirmation before generating video or long music.

Tips to avoid surprises

1. Video and long music are the big spend — your agent will confirm the approximate cost before generating them.
2. If a result isn’t perfect on the first try, your agent won’t repeat it on its own: it will ask you before retrying something expensive.
3. If you don’t have enough balance, the generation won’t happen and your agent will tell you — you can top up your balance from your client area.
4. Remember about cloned voices: persistent only on the Professional plan.

Limits worth knowing

• There’s no live voice conversation (real-time calls with your agent). What it generates are audio and video as files that arrive in your chat.
• Video clips are short (~8 seconds) and take 1–3 minutes to generate.
• The files you send it to transcribe or convert have a limit of 30 MB.