Skip to main content
Build a voice agent using Sarvam for both speech recognition and synthesis, with OpenAI GPT-4o for conversation. Optimized for Indian languages. Best for: Voice agents serving Indian language speakers with high-accuracy recognition and natural Indian voices.

​
Prerequisites

ServiceWhat You Need
PlivoAuth ID, Auth Token, Voice-enabled phone number
Sarvam AIAPI key from console.sarvam.ai
OpenAIAPI key from platform.openai.com

​
Installation

pip install "pipecat-ai[sarvam,openai]"

​
Environment Variables

# Plivo credentials
PLIVO_AUTH_ID=your_auth_id
PLIVO_AUTH_TOKEN=your_auth_token
PLIVO_PHONE_NUMBER=+1234567890

# AI service credentials
SARVAM_API_KEY=your_sarvam_key
OPENAI_API_KEY=sk-your_openai_key

​
Pipeline Configuration

from pipecat.services.sarvam import SarvamSTTService, SarvamTTSService
from pipecat.services.openai import OpenAILLMService

# Speech-to-Text (Indian languages)
stt = SarvamSTTService(
    api_key=os.getenv("SARVAM_API_KEY"),
    # model="saarika:v2",  # Saarika for STT
)

# Language Model
llm = OpenAILLMService(
    api_key=os.getenv("OPENAI_API_KEY"),
    model="gpt-4o",
)

# Text-to-Speech (Indian voices)
tts = SarvamTTSService(
    api_key=os.getenv("SARVAM_API_KEY"),
    # voice_id="your_preferred_voice",
)

​
Service Details

​
Sarvam STT

Real-time speech recognition optimized for Indian languages via WebSocket streaming.
FeatureDescription
Indian language supportHindi, Tamil, Telugu, Kannada, Malayalam, and more
Voice Activity DetectionAutomatic speech endpoint detection
Multiple audio formatsFlexible input handling
ModelsSaarika (STT), Saaras (STT with translation)

​
OpenAI LLM

Chat completion with GPT-4o supporting streaming responses and function calling.
ModelDescription
gpt-4oMost capable, multimodal
gpt-4o-miniFaster, cost-effective

​
Sarvam TTS

Text-to-speech synthesis specialized for Indian languages and voices.
FeatureDescription
Indian voicesNatural-sounding voices for multiple Indian languages
Voice customizationPitch, pace, and loudness controls
Mixed-language supportHandle content combining multiple languages

​
Supported Indian Languages

Sarvam supports multiple Indian languages including:
  • Hindi
  • Tamil
  • Telugu
  • Kannada
  • Malayalam
  • Bengali
  • Marathi
  • Gujarati
  • And more

​
Quick Start

​
Inbound Calls

git clone https://github.com/pipecat-ai/pipecat-examples.git
cd pipecat-examples/plivo-chatbot/inbound

# Configure environment
cp env.example .env
# Edit .env with Plivo, Sarvam, and OpenAI credentials

# Modify bot.py to use Sarvam services
# Start server
uv sync && uv run server.py

# Expose with ngrok (development)
ngrok http 7860
Configure your Plivo number’s Answer URL to your ngrok URL.

​
Outbound Calls

cd pipecat-examples/plivo-chatbot/outbound

cp env.example .env
uv sync && uv run server.py

# Initiate a call
curl -X POST http://localhost:7860/start \
  -H "Content-Type: application/json" \
  -d '{"phone_number": "+91XXXXXXXXXX"}'

​
When to Use Sarvam

Choose Sarvam when:
  • Your users speak Indian languages
  • You need high accuracy for Indian accents and dialects
  • You want natural-sounding Indian voices
  • Your content mixes English with Indian languages
Choose Deepgram + other TTS when:
  • Your users primarily speak English or European languages
  • You need broader language coverage
  • You’re not targeting Indian markets