> 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 full documentation content in one file, see https://docs.sarvam.ai/llms-full.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.sarvam.ai/_mcp/server.

# Introduction to conversational agents

A conversational agent is software that talks with your end customers in natural language to get something done, answer a question, collect information, complete a transaction, or hand off to a human. Instead of navigating a menu or filling a form, the customer simply speaks or types, and the agent responds.

Voice Agents is built to run these agents **across India's languages**, on the channels your customers already use.

## Voice and chat

Agents operate in one of two primary modalities:

| Modality  | The customer…      | Typical latency budget   | Notes                                                                             |
| --------- | ------------------ | ------------------------ | --------------------------------------------------------------------------------- |
| **Voice** | Speaks and listens | Sub-second, turn-by-turn | Runs a real-time ASR → LLM → TTS loop (see [Models](/conversations/build/models)) |
| **Chat**  | Types and reads    | Relaxed                  | Text-in, text-out; supports richer message types                                  |

The same underlying agent design, prompts, tools, knowledge, variables, powers both. What changes is how the customer sends and receives turns.

## Channels of deployment

An agent is authored once and can be deployed to multiple channels:

* **Telephony**: inbound and outbound phone calls (see [Telephony](/conversations/deploy/telephony))
* **Website / embedded widget**: a voice or chat widget on your web app (see [Deploy with code](/conversations/deploy/deploy-with-code))
* **WhatsApp**: chat over WhatsApp
* **API & SDK**: drive the agent directly from your own application (see [Deploy with code](/conversations/deploy/deploy-with-code))

## Where to go next

#### [Models](/conversations/build/models)

The constellation of models behind a voice agent.

#### [The harness](/conversations/build/concepts/harness)

Context, tools, and the agentic loop that wraps the models.