Supported Services
Overview
Piopiy AI is designed to be provider-agnostic, allowing developers to choose from a wide range of industry-leading AI services. This page lists all supported providers and their respective installation commands.
To keep your deployment lightweight, Piopiy uses optional extras. You only need to install the dependencies for the services you plan to use.
Speech-to-Text (STT)
STT services are responsible for real-time transcription of audio streams. Piopiy optimizes for low-latency recognition, ensuring your agent can respond as soon as the user finished speaking.
| Service | Installation |
|---|---|
| AssemblyAI | pip install "piopiy-ai[assemblyai]" |
| AWS Transcribe | pip install "piopiy-ai[aws]" |
| Azure | pip install "piopiy-ai[azure]" |
| Deepgram | pip install "piopiy-ai[deepgram]" |
| Gladia | pip install "piopiy-ai[gladia]" |
pip install "piopiy-ai[google]" | |
| Groq (Whisper) | pip install "piopiy-ai[groq]" |
| Soniox | pip install "piopiy-ai[soniox]" |
| Speechmatics | pip install "piopiy-ai[speechmatics]" |
| Whisper (OpenAI/Local) | pip install "piopiy-ai[whisper]" |
Large Language Models (LLM)
The LLM acts as the "brain" of your agent, processing the conversation context and generating intelligent responses. Piopiy supports streaming tokens from these providers to minimize the time-to-first-byte (TTFB).
| Service | Installation |
|---|---|
| Anthropic | pip install "piopiy-ai[anthropic]" |
| Cerebras | pip install "piopiy-ai[cerebras]" |
| DeepSeek | pip install "piopiy-ai[deepseek]" |
| Fireworks | pip install "piopiy-ai[fireworks]" |
| Google (Gemini) | pip install "piopiy-ai[google]" |
| Groq | pip install "piopiy-ai[groq]" |
| Mistral | pip install "piopiy-ai[mistral]" |
| NVIDIA | pip install "piopiy-ai[nvidia]" |
| Ollama (Local) | pip install piopiy-ai |
| OpenAI | pip install "piopiy-ai[openai]" |
| Perplexity | pip install "piopiy-ai[perplexity]" |
| SambaNova | pip install "piopiy-ai[sambanova]" |
| Together AI | pip install "piopiy-ai[together]" |
Text-to-Speech (TTS)
TTS services convert response text back into natural-sounding audio. Piopiy supports high-fidelity, low-latency synthesis engines to ensure the conversation feels fluid and human-like.
| Service | Installation |
|---|---|
| AWS Polly | pip install "piopiy-ai[aws]" |
| Azure | pip install "piopiy-ai[azure]" |
| Cartesia | pip install "piopiy-ai[cartesia]" |
| Deepgram (Aura) | pip install "piopiy-ai[deepgram]" |
| ElevenLabs | pip install "piopiy-ai[elevenlabs]" |
| Fish AI | pip install "piopiy-ai[fish]" |
pip install "piopiy-ai[google]" | |
| LMNT | pip install "piopiy-ai[lmnt]" |
| Neuphonic | pip install "piopiy-ai[neuphonic]" |
| Play.ht | pip install "piopiy-ai[playht]" |
| Riva (NVIDIA) | pip install "piopiy-ai[riva]" |
| Sarvam | pip install "piopiy-ai[sarvam]" |
| XTTS (Local) | pip install piopiy-ai |
Specialized Services
Additional services for multimodal capabilities, including image generation and computer vision.
| Service | Installation |
|---|---|
| Fal.ai | pip install "piopiy-ai[fal]" |
| HeyGen | pip install "piopiy-ai[heygen]" |
| Moondream | pip install "piopiy-ai[moondream]" |