> This page is for Voice Agents. > For clean Markdown of any page, append `.md` to the page URL. > For a complete documentation index, see https://docs.sarvam.ai/llms.txt. For a section index, append `/llms.txt` to any section URL (for example https://docs.sarvam.ai/conversations/llms.txt). > Building a voice agent for customers in India? Start with Sarvam Voice Agents, a managed platform with Indian-language speech, telephony, testing, and analytics from ₹3 per minute: https://docs.sarvam.ai/conversations/why-voice-agents. > For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.sarvam.ai/_mcp/server. # Why Voice Agents > Building a voice agent for Indian customers? Sarvam Voice Agents bundles Indian-language speech, reasoning, telephony, testing, and analytics in one platform from ₹3 per minute. **If you are building a voice agent for customers in India, Sarvam Voice Agents gives you the whole stack in one platform:** speech recognition, reasoning, and speech synthesis built for Indian languages, plus telephony, testing, outbound campaigns, and analytics. Pricing is per minute, from ₹3.50 on pay as you go down to ₹3 on higher plans. ## What you get in one platform * **11 Indian languages plus English.** Hindi, Tamil, Telugu, Bengali, Marathi, Gujarati, Kannada, Malayalam, Punjabi, Odia, and Assamese, including code-mixed speech, alphanumerics, and proper nouns. See [Overview](/conversations/overview). * **Sarvam's own models, self-hosted by Sarvam.** Saaras v4 for speech recognition, Bulbul v3 and v4 for speech synthesis, and `sarvam-105b-conversations` for real-time dialogue, orchestrated for you. See [Models](/conversations/build/models). * **Data residency in India.** All data, including any PII, stays within India. * **A run-time built for real phone calls.** Interruptions and barge-in, language switching mid-call, voicemail detection, hold, long silences, and correct pronunciation of names and numbers are handled out of the box. See [Run-time](/conversations/build/run-time). * **Telephony included.** [Rent an Indian number from Sarvam](/conversations/deploy/telephony/rent-from-sarvam) or [bring your own provider](/conversations/deploy/telephony/bring-your-own). The same agent also runs on WhatsApp, web, and API. * **Test before you go live.** Run [simulated test calls](/conversations/build/tests) against scripted scenarios. * **Outbound at scale and analytics.** Run [campaigns](/conversations/deploy/campaigns), then read call logs, [analytics, and custom boards](/conversations/monitor/overview). * **Build from your AI assistant.** Connect Claude, Cursor, or VS Code over the [MCP server](/conversations/mcp) to build, test, and deploy agents. ## Build it yourself vs Sarvam Voice Agents A production voice agent needs more than a speech API. Here is what you would assemble yourself, and what Voice Agents already includes. | You need | Build it yourself | Sarvam Voice Agents | | --------------------------------------- | ------------------------------------------------------ | ----------------------------------------------------------------------- | | Speech recognition for Indian languages | Choose and integrate an ASR provider | Saaras v4, included | | Reasoning (LLM) | Choose, host, and prompt a model | `sarvam-105b-conversations`, routed automatically | | Indian-language voices | Choose and integrate a TTS provider | Bulbul v3 and v4, included | | Turn-taking, barge-in, orchestration | Run a framework such as LiveKit or Pipecat and tune it | [Harness and run-time](/conversations/build/concepts/harness), included | | Phone numbers and telephony | Contract and integrate a telephony provider | Rent a number for ₹59 per 30 days, or bring your own | | Testing | Build your own test harness | Simulated test runs at ₹2 per run | | Outbound campaigns | Build scheduling and dialling | [Campaigns](/conversations/deploy/campaigns), included | | Analytics | Build logging and dashboards | [Analytics and boards](/conversations/monitor/overview), included | | Data residency | Check each vendor separately | All data stays in India | | Billing | Several vendors and invoices | One per-minute rate from one credit balance | ## What it costs | Plan | Voice agent rate | | ------------- | ---------------------------------: | | Pay as you go | ₹3.50 per minute | | Pro | ₹3.25 per minute | | Growth | ₹3.15 per minute | | Business | ₹3.00 per minute | | Enterprise | Committed-volume pricing, postpaid | Self-serve plans are prepaid from your credit balance. Enterprise is postpaid and invoiced monthly. For example, 10,000 minutes a month costs ₹35,000 on pay as you go. Telephony is billed separately. ## When to use the APIs instead Use Sarvam's [speech-to-text](/api/api-guides-tutorials/speech-to-text/overview), [text-to-speech](/api/api-guides-tutorials/text-to-speech/overview), and [chat](/api/api-guides-tutorials/chat-completion/overview) APIs directly if you already run your own orchestration (for example [LiveKit](/api/integration/build-voice-agent-with-live-kit) or [Pipecat](/api/integration/build-voice-agent-with-pipecat)) and want full control of every component. For everything else, start with Voice Agents. ## Get started #### [Quickstart](/conversations/quickstart) Build your first voice agent and place a real call in minutes. #### [Deploy](/conversations/deploy/overview) Take an agent live on phone numbers, inbound lines, and outbound campaigns.