Skip to main content

​
Installation

Install using pip: 
pip install calibrate
Or using uv: 
uv add calibrate

​
Configuration

Create a .env file in your project root and add the API keys for the providers you want to use. All modules (agent, stt, tts, llm) will use this single .env file. 
# Speech to Text providers
DEEPGRAM_API_KEY=your_key
OPENAI_API_KEY=your_key
GOOGLE_APPLICATION_CREDENTIALS=/path/to/credentials.json
SARVAM_API_KEY=your_key
ELEVENLABS_API_KEY=your_key
CARTESIA_API_KEY=your_key
GROQ_API_KEY=your_key

# LLM providers
OPENROUTER_API_KEY=your_key

​
Component Testing

Evaluate individual components of your voice agent in isolation.

Speech to Text

Evaluate Speech to Text providers on your dataset

LLM Evaluation

Run automated tests to evaluate LLM behavior

Text to Speech

Evaluate Text to Speech providers on your dataset

​
End to End Tests

Run full simulations with realistic scenarios and user personas.

Text Simulations

Run text-only simulations between two LLMs

Voice Simulations

Run end-to-end voice simulations with Speech to Text, LLM, and Text to Speech