Skip to content
DreamBuilds

Best for voice apps

Best AI tools for building voice apps

A voice app usually needs three pieces: a way to hear the user (speech-to-text), a way to answer out loud (text-to-speech), and sometimes an agent that handles the whole call. These listings cover each piece.

6 tools. Ordered by how directly each listing matches this job, judged from its DreamBuilds description. This is an editorial order, not a benchmark or a test result.

  1. Vapi

    API for voice agents that place and answer phone calls.

    Handles the whole phone call as a voice agent.

    Official site

  2. ElevenLabs

    Speech synthesis and voice cloning, plus tools for voice agents.

    Official site

  3. Deepgram

    Speech-to-text API for recorded audio and live streams.

    Speech-to-text for both recorded audio and live streams.

    Official site

  4. Cartesia

    Text-to-speech API for low-latency spoken audio.

    Text-to-speech described as low-latency.

    Official site

  5. AssemblyAI

    Speech-to-text API for transcription.

    Official site

  6. Inworld

    Character engine for games, including spoken dialogue.

    Spoken dialogue for game characters.

    Official site

FAQ

Which AI tools turn speech into text?
DreamBuilds lists Deepgram (recorded audio and live streams) and AssemblyAI (transcription) as speech-to-text APIs.
Which tools generate spoken audio?
ElevenLabs (speech synthesis and voice cloning) and Cartesia (low-latency text-to-speech) are listed for spoken output.
Can I build an AI phone agent?
Vapi is listed as an API for voice agents that place and answer phone calls.

More lists

Built an AI tool? Get it listed.