> Sarvam Voice Agents (Samvaad) is a voice AI agent platform for Indian languages: build, test and deploy AI calling agents on your own phone numbers, from ₹3 a minute. Start here: https://docs.sarvam.ai/voice-agents/why-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/voice-agents/llms.txt). > For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.sarvam.ai/_mcp/server. # How-to ## Docs - [Choose the Script of the Text](https://docs.sarvam.ai/api/api-guides-tutorials/dictate/how-to/choose-script.md): Use mode to pick the script of the text Dictate returns: Hinglish, Roman Hindi, one script, or every word as spoken. - [Get Names Right](https://docs.sarvam.ai/api/api-guides-tutorials/dictate/how-to/custom-vocabulary.md): Make product, customer and medicine names come out with your spelling, using vocabulary and context.reference. - [Change the Form of the Text](https://docs.sarvam.ai/api/api-guides-tutorials/dictate/how-to/style-rules.md): Use style.preset and style.rules to drop fillers, make bullets or checklists, translate, and write rupees and times your way. - [Get JSON, Summaries and Answers](https://docs.sarvam.ai/api/api-guides-tutorials/dictate/how-to/structured-output.md): Use output_rules to turn the whole recording into JSON for a form or CRM, a summary, answers to set questions, or a reply draft. - [Transcribe Long Recordings](https://docs.sarvam.ai/api/api-guides-tutorials/dictate/how-to/long-recordings.md): Use a Dictate job for a recording up to 1 hour: upload or pass a URL, poll or get a callback, read the result, then delete the job. - [Dictate Live](https://docs.sarvam.ai/api/api-guides-tutorials/dictate/how-to/live-dictation.md): Show words while the user speaks, with one finished text after each pause, over the Dictate WebSocket. Python and JavaScript SDK examples. - [Pin a Model Version](https://docs.sarvam.ai/api/api-guides-tutorials/dictate/how-to/pin-model-version.md): Pin a dated Dictate snapshot with the model setting, so text that your code parses does not change without notice. - [Handle Errors and Retries](https://docs.sarvam.ai/api/api-guides-tutorials/dictate/how-to/error-handling.md): How the Dictate SDKs raise errors, which errors to retry, and a try/except sample in Python and JavaScript. - [Audio and Languages](https://docs.sarvam.ai/api/api-guides-tutorials/dictate/how-to/audio-and-languages.md): The audio formats, sizes and lengths Dictate accepts for a clip, a job and the live socket, and the 23 supported language codes.