AI agents
AI agents from ₹1 a minute.¹
Speech recognition, the model and the voice, in one rupee rate. Billed in 10-second blocks, so short calls stay cheap.
Choose your planLiveKit alternative
LiveKit gives you the pipes. We give you the agent, the models and the phone line, in one rupee price.
AI agents from ₹1 a minute.¹
Why switch
Everything a phone call in India needs, from one team, priced in rupees.
AI agents from ₹1 a minute.¹
Speech recognition, the model and the voice, in one rupee rate. Billed in 10-second blocks, so short calls stay cheap.
Choose your planHindi, English, Gujarati, Tamil and Telugu. And the Hinglish in between, mid-sentence.
The listening, the thinking and the voice are ours: Fourier STT, Gemma 4 26B and Fourier TTS.
The phone line, the platform, speech recognition, the model and the voice all come from us.
Build in the sandbox, press Go live, and call today. Every company starts on pay as you go, with no monthly commitment.
Real lines from Fourier TTS, one request each, unedited. Pick a language and press play.
Hear itYou build the agent, or you buy the call
LiveKit gives developers the plumbing: an open-source agent framework, WebRTC and SIP, and hosted models you pay for by the minute. You write and run the agent. We are the other side of that choice: speech recognition, model and voice of our own, on our GPUs in India, sold as one rupee price per minute with the phone line inside on pay as you go.
LiveKit is what a good engineering team reaches for when it wants to own the whole thing. The agent framework is open source. WebRTC and SIP are in one stack. Models come from Deepgram, OpenAI, Cartesia and others, hosted by LiveKit or on your own keys. You write the agent, pick each part, and deploy it.
That is a real freedom, and it comes with a real to-do list. Someone on your team writes and hosts the code. Someone chooses the Hindi recogniser and the Hindi voice and checks they work together. Someone sets up the SIP trunk to an Indian carrier, because LiveKit’s own numbers are US-only and inbound-only. And the bill arrives as several lines.
We sell the whole call. Speech recognition, the model and the voice are ours. The model is the open Gemma 4 26B, on our own GPUs in India, where every call runs. You write a prompt, pick a voice, and the agent is live the same day. On pay as you go the phone line is in the price.
Real lines from Fourier TTS. Pick one, press play.
Text149 characters
Hi, मैं नेहा बोल रही हूँ Acme Store से। आपका order dispatch हो गया है और Friday तक deliver हो जाएगा। Address में कोई change करना हो तो अभी बता दीजिए।
48 kHz · MonoReady when you are
0:00 / 0:09
Fourier TTS, voice neha, rendered 16 Oct 2026. One request, not edited.
Text119 characters
नमस्ते, मैं श्रुति बोल रही हूँ, दाँतों के क्लिनिक से। कल सुबह डॉक्टर साहब से आपका मिलना तय है। क्या आप समय पर आ पाएँगे?
48 kHz · MonoReady when you are
0:00 / 0:08
Fourier TTS, voice shruti, rendered 16 Oct 2026. One request, not edited.
Text117 characters
નમસ્તે, હું આશુ બોલું છું. તમારી લોનનો હપ્તો આવતા અઠવાડિયે ભરવાનો છે. તમે કહો તો હું અત્યારે જ ભરવાની રીત સમજાવી દઉં.
48 kHz · MonoReady when you are
0:00 / 0:08
Fourier TTS, voice ashu, rendered 17 Oct 2026. One request, not edited.
Text167 characters
வணக்கம், நான் நேஹா பேசுகிறேன். உங்கள் பார்சல் நாளை மாலைக்குள் வந்து சேரும். அந்த நேரத்தில் நீங்கள் வீட்டில் இல்லையென்றால் சொல்லுங்கள், வேறு நாளுக்கு மாற்றி விடுகிறோம்.
48 kHz · MonoReady when you are
0:00 / 0:10
Fourier TTS, voice neha, rendered 17 Oct 2026. One request, not edited.
Text114 characters
నమస్కారం, నేను సమీర్ మాట్లాడుతున్నాను. నిన్న మీరు మా కొత్త ఇళ్ళ గురించి అడిగారు. ఇప్పుడు కొంచెం సేపు మాట్లాడవచ్చా?
48 kHz · MonoReady when you are
0:00 / 0:08
Fourier TTS, voice sameer, rendered 17 Oct 2026. One request, not edited.
Side by side
| Option | Bar | ₹ a minute |
|---|---|---|
| Pay as you goLiveKit Cloud's Ship plan, against our T5 | ||
| ₹0.95$0.01 | ||
| ₹4.55$0.0479 | ||
| ₹5all-in | ||
| At volumeLiveKit Cloud's Scale plan, against our T3 and T1 | ||
| ₹0.95$0.01 | ||
| ₹3all-in | ||
| ₹1₹1.50 with both | ||
Dollar prices converted at ₹94.96, the ECB reference rate on 16 October 2026. The rate.
Each bar is one minute as LiveKit’s own pages price it, layer by layer, beside ours from our price list. A hatched end is a part of the call their price leaves out and publishes no number for, so it has no length.
LiveKit's pricing page adds up a phone call on Build or Ship as the agent session, its US telephony, the language model, speech to text, text to speech and observability: $0.0479 a minute. The line in it is LiveKit's own US inbound number, so a call to India needs a carrier's trunk instead, at the carrier's rate plus LiveKit's third-party SIP minutes. The Ship and Scale plans also carry a monthly fee, $50 and $500, with minutes included.
From livekit.com/pricing, checked . Ours from our price list.
The chart shows the plan you choose above beside the price it pairs with: our T1 and T3 beside LiveKit’s volume price, our T5 beside its pay-as-you-go price. Without that choice, every bar is listed.
| Detail | LiveKit | Frequency |
|---|---|---|
| Published price per minute | Agent sessions: Build $0/mo with 1,000 minutes included Ship $50/mo with 5,000 minutes included then $0.01 per min; Scale $500/mo with 50,000 minutes included then $0.01 per min; the models and the phone line are billed on top livekit.com | T1 ₹1/min with ₹1,00,000/month T3 ₹3/min with ₹50,000/month; T5 ₹5/min pay as you go, phone line and insights included (GST extra) |
| What else a minute adds | Models through LiveKit Inference at each model's price, for example Deepgram Nova-3 (Multilingual) at $0.0058 a minute and Cartesia Sonic 3 at $50.00 per million characters on Build and Ship third-party SIP minutes at $0.004 per min on Ship and $0.003 per min on Scale; and your carrier's own charges for the call livekit.com | Nothing on T5 and T3: the phone line and insights are inside the rate on T1 they are optional at ₹0.30/min and ₹0.20/min |
| Their worked example | LiveKit's own example of a phone call on Build or Ship comes to $0.0479 a minute: the agent session, its US telephony, the model, speech recognition, the voice and observability a carrier's line to India is not in it livekit.com | 1,000 minutes on T5: 1,000 × ₹5 = ₹5,000, phone line included |
LiveKit’s pricing page does the sum for you. A phone call on its Build or Ship plan comes to $0.0479 a minute with its hosted models and a US line, about ₹4.55 at the ECB rate. The agent session in that is $0.01 per min, about ₹0.95. The Ship plan adds $50 a month, ₹4,748, with 5,000 session minutes inside. A call to India swaps the US line for a carrier’s trunk, at the carrier’s rate plus LiveKit’s third-party SIP minutes.
Ours is ₹5 a minute on pay as you go, line included, so 1,000 minutes cost ₹5,000. When you’re calling every day, the volume plan takes it to ₹1 a minute. We bill in 10-second blocks, and a call nobody answers is free. The build-or-buy guide walks through the trade.
LiveKit is infrastructure: it carries the call and runs your code, and the models come from other companies at their price. We are the models and the platform.
| Detail | LiveKit | Frequency |
|---|---|---|
| Phone line | You bring it Twilio, Telnyx, Exotel, Plivo or another SIP trunk docs.livekit.io | Ours Mirai phone line, +91 numbers |
| Platform | Theirs LiveKit Cloud livekit.com | Ours Our orchestration |
| Speech recognition | Third party, your pick Deepgram, AssemblyAI, Cartesia and others, hosted or on your keys livekit.com | Ours Fourier STT |
| Language model | Third party, your pick OpenAI, Google and others livekit.com | Ours Gemma 4 26B, on our GPUs |
| Voice | Third party, your pick Cartesia, Deepgram and others; ElevenLabs and Sarvam as plugins docs.livekit.io | Ours Fourier TTS |
| Detail | LiveKit | Frequency |
|---|---|---|
| Languages | Whatever the model you pick supports Hindi is in the Inference lists for Deepgram Nova-3, AssemblyAI and Cartesia Sonic 3, and Sarvam is a plugin on your own key, not in Inference docs.livekit.io | Hindi, Gujarati, English, Tamil and Telugu, plus Hinglish, in speech recognition and voice Bengali, Marathi and Kannada are coming |
| Detail | LiveKit | Frequency |
|---|---|---|
| Where it runs | A global edge network with region pinning on Scale SOC 2 Type II; PCI DSS and ISO 27001 are in progress; no India region is named on the pages we read livekit.com | Every call runs on our own GPUs in India |
| Detail | LiveKit | Frequency |
|---|---|---|
| Phone numbers | LiveKit Phone Numbers are available only in the US, and only inbound calling is supported calls to India go through a SIP trunk from Twilio, Telnyx, Exotel, Plivo or another carrier docs.livekit.io | Indian (+91) numbers: buy a dedicated number through us, or connect your own Plivo number |
| Detail | LiveKit | Frequency |
|---|---|---|
| Concurrent agent sessions | 5 on Build, 20 on Ship, Up to 600 on Scale, custom on Enterprise livekit.com | A new workspace starts with 5 calls at a time, raised when you ask, with no code change |
| Detail | LiveKit | Frequency |
|---|---|---|
| Your own code | Open source under Apache 2.0 you write the agent in Python or Node.js and host it on LiveKit Cloud or your own servers docs.livekit.io | Prompt or node-based agents built in the console, started from your code with POST /v2/calls |
Spot something out of date? Tell us. Six platforms in one table: Voice AI platforms compared.
Which to pick
Sometimes it's them. We'll say so.
If you want to self-host, LiveKit is open source and we aren’t. If you want Sarvam’s recogniser with ElevenLabs’s voice and your own fine-tuned model, LiveKit will wire them up. It hands calls to a person, cold or warm, which we can’t do yet. Its Scale plan runs up to 600 agent sessions at once; a new workspace with us starts at 5 and grows when you ask. And it carries SOC 2 Type II today.
Moving
Bring the agent you have. Most of it carries across.
Step 1
Look around with no sign-up. We ask for an email only when you test a call.
Step 2
Paste the prompt you already wrote into Agent Studio, and pick a voice and a language.
Step 3
Test it in your browser or on your own phone, then go live the same day.
Your system prompt becomes the agent’s prompt in Agent Studio, and your first message its greeting. The tool calls you wrote become the agent’s tools, and the result of every call arrives as a signed webhook. If your agent follows a fixed script, lay it out as nodes instead. The quickstart goes from a new key to a real phone call in five minutes, and the latency guide shows how to time the reply on both.
Questions
LiveKit's own example prices a minute at $0.0479 with its hosted models and a US line, so 1,000 minutes come to about $47.90, or ₹4,549 at the ECB rate, before a carrier's line to India and the Ship plan's $50 a month. On our pay-as-you-go plan, 1,000 minutes cost ₹5,000 with the line included, and the agent is already written.
No. The $0.01 per min, about ₹0.95, is the agent session alone. The speech recognition, the model, the voice and the phone line are billed on top, by LiveKit Inference, by your own providers or by your carrier. Our ₹5 on pay as you go is all of them together.
On LiveKit, yes. You write the agent in Python or Node.js, choose the models and host it. With us, you write a prompt in the console, pick a voice and a language, and test it in your browser. Your own code starts calls with one POST and reads the result from a webhook.
Fourier TTS speaks OpenAI's POST /v1/audio/speech, so a client that lets you set a base URL and a model name can call it. Our Pipecat guide shows the same idea with pipecat-mirai. Custom setups are on you to test.
Fifteen minutes with the team that builds it. Bring your script, or your hardest call.
Book 15 minutes