---
title: "Voice AI Providers Supported by Micdrop | Micdrop"
description: "Every AI provider Micdrop integrates with, and the parts of a voice agent each one covers: speech to text, text to speech, LLM, realtime speech to speech and turn classification."
url: "https://micdrop.dev/docs/providers"
---

# Voice AI Providers Supported by Micdrop

A Micdrop server hands each part of the call to a provider of your choice, so you can mix them for cost and quality, run them all locally, or write your own.

*   **[Speech to text](/docs/speech-to-text)** transcribes the user.
*   **[LLM (Agent)](/docs/agents)** writes the answers.
*   **[Text to speech](/docs/text-to-speech)** speaks them.
*   **[Realtime](/docs/realtime)** models hear and speak in one connection, in place of the three above.
*   **[Classifiers](/docs/classifiers)** turn each turn of the user into typed decisions next to the agent.

Each provider ships as its own npm package. The table shows which part of a voice agent each one covers, with a link to its page.

Provider

Package

Speech to text

Text to speech

LLM

Realtime

Classifier

Vercel AI SDK

`@micdrop/ai-sdk`

[Yes](/docs/agents/ai-sdk)

Cartesia

`@micdrop/cartesia`

[Yes](/docs/text-to-speech/cartesia)

ElevenLabs

`@micdrop/elevenlabs`

[Yes](/docs/text-to-speech/elevenlabs)

Gemini

`@micdrop/gemini`

[Yes](/docs/speech-to-text/gemini)

[Yes](/docs/text-to-speech/gemini)

[Yes](/docs/agents/gemini)

[Gemini Live](/docs/realtime/gemini-live)

[Decisions](/docs/classifiers/gemini-decisions)

Gladia

`@micdrop/gladia`

[Yes](/docs/speech-to-text/gladia)

Gradium

`@micdrop/gradium`

[Yes](/docs/speech-to-text/gradium)

[Yes](/docs/text-to-speech/gradium)

Kokoro (local)

`@micdrop/kokoro`

[Yes](/docs/text-to-speech/kokoro)

Mistral

`@micdrop/mistral`

[Voxtral](/docs/speech-to-text/mistral)

[Voxtral](/docs/text-to-speech/mistral)

[Yes](/docs/agents/mistral)

OpenAI

`@micdrop/openai`

[Yes](/docs/speech-to-text/openai)

[Yes](/docs/text-to-speech/openai)

[Yes](/docs/agents/openai)

[Realtime](/docs/realtime/openai-realtime)

[Decisions](/docs/classifiers/openai-decisions)

Phonon (local)

`@micdrop/phonon`

[Yes](/docs/text-to-speech/phonon)

Piper (local)

`@micdrop/piper`

[Yes](/docs/text-to-speech/piper)

Pocket TTS (local)

`@micdrop/pocket-tts`

[Yes](/docs/text-to-speech/pocket-tts)

Qwen3-TTS (local)

`@micdrop/qwen-tts`

[Yes](/docs/text-to-speech/qwen-tts)

TypeSafe

`@micdrop/typesafe`

[Jev](/docs/classifiers/jev)

Whisper (local)

`@micdrop/whisper`

[Yes](/docs/speech-to-text/whisper)

The AI SDK package opens the LLM column to every model the [Vercel AI SDK](/docs/agents/ai-sdk) supports, Anthropic, Groq or a local Ollama server included. Any other provider plugs in through a [custom speech to text](/docs/speech-to-text/custom), [text to speech](/docs/text-to-speech/custom), [agent](/docs/agents/custom) or [classifier](/docs/classifiers/custom).

Mistral, Gladia and Gradium cover the three parts of a call with European companies, as the [sovereign voice AI guide](/docs/providers/sovereign-voice-ai) shows.

## Mixing providers

Each part is a separate option of the server, so every one can come from a different provider:

```typescript
import { MicdropServer } from '@micdrop/server'
import { MistralAgent } from '@micdrop/mistral'
import { GladiaSTT } from '@micdrop/gladia'
import { CartesiaTTS } from '@micdrop/cartesia'

new MicdropServer(socket, {
  stt: new GladiaSTT({
    apiKey: process.env.GLADIA_API_KEY,
  }),
  agent: new MistralAgent({
    apiKey: process.env.MISTRAL_API_KEY,
    model: 'mistral-large-latest',
  }),
  tts: new CartesiaTTS({
    apiKey: process.env.CARTESIA_API_KEY,
    voiceId: 'cartesia-voice-id',
  }),
})
```

The speech to text, the agent and the text to speech all have a local counterpart, so a whole conversation can also run on your machine with the [local models guide](/docs/local-models).
