Crackle — AI voice agents for your website and app

Write what your agent should do, pick a voice, and talk to it in your browser. When it sounds right, put it on your website with one line of HTML, or start calls from your app with one request.

Start free See how it works

Free to start, with sign-up credit for your first test calls.

Sample call0:00

Taraconnecting

One agent, 40+ languages, each written in its own script before it is spoken.

हिन्दी
English
தமிழ்
Español
తెలుగు
Français
বাংলা
日本語
मराठी
Deutsch
ગુજરાતી
العربية
ಕನ್ನಡ
Português
മലയാളം
한국어

ਪੰਜਾਬੀ
Italiano
اردو
中文
Türkçe
Русский
Tiếng Việt
ไทย
Bahasa Indonesia
Nederlands
עברית
Polski
Ελληνικά
Svenska
Українська
Tagalog

The feel

The small things that make a call feel human.

Timing, interruptions, tone. The parts of a conversation nobody notices until they go wrong.

Cut in any time

Start talking and the agent stops mid-word and listens, the way a person would. It picks the thread back up when you are done.

Not thrown by a “हाँ”

A quick “okay” or “mm-hm” while it is talking is a listener nodding along, not an interruption. It keeps going.

Laughs, sighs, slows down

Set the pace and the tone per agent. On voices that can perform them, a written laugh or sigh is acted out, never read aloud.

Follows the caller’s language

When a caller moves between Hindi and English, the agent moves with them and stays there for the rest of the call.

Knows when it’s over

Once the caller is done, it says one short goodbye and hangs up, instead of leaving them on an open line.

How it works

From a blank page to a live call in four moves.

Brief it like a new hire

An agent’s instructions are a page of plain writing, not a flowchart. Say who it is, what it is for and what it must never do.

Variables fill in what you know about each caller
Tools are your own HTTP endpoints, used mid-conversation
A greeting it speaks first, written or improvised

Tara Saved Test agent

Greeting

Hi customer_name, this is Tara at Sunrise Dental. How can I help?

Instructions

You answer calls for a dental clinic and look after its bookings. Use find_booking to look up the caller’s appointment and reschedule to move it.

Give it a voice

Pick the speech engine, the voice and the language, then the model that does the thinking. Sliders set the pace and the volume.

Search the voice library for the right voice
One price per minute, whichever voice and model you pick
Press Test agent and talk to it in your browser

Voice engineWhat turns the reply into speech

Cartesia SonicRumik Mulberry

LanguageWritten in its own script

हिन्दी · Hindi

SpeedHow fast it talks

1.00×

ToneThe mood it speaks in

CalmContentNeutral

ModelWhat does the thinking

Gemma 4 31B

Every combination costs the same per minute.

Ship it with one request

Every agent has its own API key. One POST starts a call and returns what a LiveKit client needs to join it, from a browser or a phone.

Per-call variables and your own metadata
Signed webhooks, retried if your server is down
Live captions for your own call screen

POST /v1/calls

curl -X POST "$CRACKLE_API_URL/v1/calls" \
  -H "Authorization: Bearer $CRACKLE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "participant": { "identity": "customer_123" },
    "variables": { "customer_name": "Divya" }
  }'

Returns

{ "callId": "…", "livekitUrl": "wss://…", "token": "eyJ…" }

Then your webhook is told

call.started: The call exists and an agent is on its way
call.user_joined: Your caller is on the line
transcript.user: Each thing the caller finished saying
transcript.assistant: Each complete reply
call.ended: Why it ended, how long it ran, the full transcript

Read every call back

Each call is kept with its transcript, how long it ran and what it cost, next to the numbers that say how it went.

Who cut in, and how often
How long each reply took to start
Why the call ended

TaraCompletedsample call

Caller turns: 2
Agent replies: 3
Interruptions: 1
How it ended: The agent ended the call

Time to reply, turn by turn

CallerI need to move my appointment on Friday.

TaraGot it, Friday at four thirty with Doctor Rao. I have Monday at ten, or Tues—

CallerMonday works.

Features

Everything your agent needs to hold a conversation.

You bring what it should know and do. Crackle brings the voice, the listening, the timing and the call, already wired together.

A voice in 44 languages

Hindi, Hinglish, Tamil, English and forty more, each reply written in its own script before it is spoken.

Talks like a person

Cut in and it stops to listen. A quick “हाँ” or “okay” does not throw it, and it hangs up once the caller is done.

Does things, not just talks

Your own HTTP endpoints, called in the middle of a conversation: look up an order, book a slot, update a record.

Answers from your documents

Upload price lists, policies and FAQs. The agent answers from them, and says so when it does not know.

Knows who it is talking to

Fill the instructions with what you know about each caller: their name, their booking, their plan.

On your website in one line

Paste one line of HTML and a talk button appears on your site. No server of your own.

In your web or mobile app

Join a call from a web page, an iOS or Android app, or React Native, with live captions.

Every call on record

Transcripts, durations and costs in the call log, and sent to your server as each call happens.

Built and tested in your browser

Write the instructions, pick the voice, press Test agent and talk to it. No code until you want it.

Pricing

Free to build. Priced by the minute.

Build and test without a plan. When you go live, a plan adds call credit every month and room for more calls at once.

Free

Build and test your first agent.

$0to start

Sign-up credit for test calls
2 calls at a time
All voices and languages
Studio, tools and webhooks

Starter

For a product going live.

$29per month

$30 of call credit every month
5 calls at a time
API keys for your app
Email support

Growth

For steady call volume.

$99per month

$110 of call credit every month
20 calls at a time
Priority support
Usage and call quality reports

Scale

For teams running calls all day.

$399per month

$460 of call credit every month
75 calls at a time
Dedicated support channel
Custom voices on request

Calls are billed from your credit at $0.07 a minute, whichever voice and model your agent uses. Need more than Scale? Write to us.

Questions

Before you start.

What is Crackle?

Crackle gives your AI agent a voice. You write what the agent should do, choose a voice and talk to it in your browser. When it sounds right, you put it on your website with one line of HTML, or start calls from your own app with one API request.

Do I need to write code?

Not to build and test one. The studio does both in your browser. Putting the agent on your website is one line of HTML to paste; starting calls from your own app takes one request from your server.

Which languages can an agent speak?

44: English, Spanish, French, German, Portuguese, Italian, Japanese, Korean, Chinese and Arabic among them, and Hindi, Hinglish, Tamil, Telugu, Kannada, Marathi, Bengali, Gujarati and Punjabi with voices recorded in each. Every reply is written in its language’s own script before it is spoken, and an agent can move between Hindi and English when the caller does.

How much does it cost?

Calls are billed by the second at $0.07 a minute, whichever voice and model the agent uses. Building and testing is free, with sign-up credit for your first test calls. Paid plans start at $29 a month and include call credit.

Can the agent use my own data and systems?

Yes. Upload your price lists, policies and FAQs and it answers from them. Give it your own HTTP endpoints as tools and it can look up an order or book a slot in the middle of a call.

Where do people talk to my agent?

In your website, through a talk button you paste in, or in your web, iOS, Android or React Native app. Calls run on LiveKit, so any of its client SDKs works.

Put a voice on the line.

Build your first agent and talk to it today. Sign-up credit is there for your first test calls.

Start free Sign in

Give your AI agent a voice.