DAVE is a digital voice assistant built using OpenAI's powerful technologies including Whisper ASR API, GPT-4 Chat models, and Text-To-Speech (TTS) APIs. With DAVE, you can smoothly interact with a virtual assistant that listens to you attentively, processes the information with a remarkable understanding, and articulates a reply, all made perceivable with intuitive transition states.

As DAVE is just a static html page, some javascript, and a couple images, running and hosting it is easy. If you're a python developer you can use something like python3 -m http.server 8080, otherwise there are a variety of options from VSCode extensions like Live Server to full NGINX deployments.

A demo deployment of DAVE is additionally hosted as a static site here here using DigitalOcean's App Platform.

In order to use DAVE, an OpenAI account is required with the access to the following model endpoints.

• Transcriptions /v1/audio/transcriptions
• Chat /v1/chat/completions
• Speech /v1/audio/speech

For basic usage of DAVE, the only configuration required is for you to provide your OpenAI API key within the page url apiKey parameter.

e.g. ?apiKey=sk-SsMzb...z07l

For more advanced usage, DAVE allows the voice, chat and text-to-speech models to be configured additionally using url parameters. It is important to note that for the models, such as GPT-4, it will depend on your level of API access.

Below are the list of availible configuration options (as of 2023/11/08):

• Voice Models (URL Param 'voiceModel'):

  • alloy
  • echo
  • fable
  • onyx (default)
  • nova
  • shimmer

• Chat Models (URL Param 'chatModel'):

  • gpt-3.5-turbo (default)
  • gpt-3.5-turbo-0301
  • gpt-3.5-turbo-0613
  • gpt-3.5-turbo-1106
  • gpt-3.5-turbo-16k
  • gpt-3.5-turbo-16k-0613
  • gpt-3.5-turbo-instruct
  • gpt-3.5-turbo-instruct-0914
  • gpt-4
  • gpt-4-0314
  • gpt-4-0613
  • gpt-4-1106-preview

• Text-to-Speech Models (URL Param 'textToSpeechModel'):

  • tts-1 (default)
  • tts-1-hd
  • tts-1-1106
  • tts-1-hd-1106

• Text-to-Speech Models (URL Param 'textToSpeechModel'):

  • You are a helpful voice assistant. (default)
  • custom

• Conversation Mode (URL Param 'conversationMode'):

  • true (default)
  • false

Note: These configuration options are additionally displayed within the console log for convenient reference.

If you wish to run DAVE using docker, you can use the following command to build and deploy as a docker image.

docker build -t dave .

docker run -d \
    --name dave \
    -p 8080:80 \
    dave

• Automatically cut off recording after n seconds of silence.

The development of DAVE was inspired by the excellent work of Romain Huet, who demo'd a similar (albeit more complex) application at the 2023 OpenAI DevDay. His excellent presentation is availible to watch here on YouTube.

The code and documentation in this project are released under the GPLv3 License.