WAAS (Whisper as a Service) is a platform that offers users a graphical user interface (GUI) and an API with queuing for OpenAI Whisper. Its primary function is to provide users with a service for transcribing audio or video files to text and delivering the output as an SRT (SubRip Subtitle) or text file via email. WAAS has an editor that locates 100% locally in a user's browser, providing users a simple way to play and fix transcription errors in selected segments of text. The editor is compatible with the control key on the keyboard, making it a seamless process.

WAAS has an API with numerous features to make the transcribing process smoother. Its POST /v1/transcribe endpoint allows users to add new transcribe jobs to the queue. The job is asynchronous and processed by the worker. The response is a JSON object with job_id used to check the job status. The endpoint requires a binary image of the audio content to transcribe, while other optional parameters include language, model, filename, and task. Other endpoints include POST /v1/detect for detecting the language of an audio file and GET /v1/download/<job_id> for receiving the output result, among others.

WAAS is highly compatible and requires a minimal VRAM or GPU (graphics processing unit) to function optimally. It operates within the range of Python 3.8-3.10 and is available as a remote development extension (for containers), among others. WAAS has robust documentation to guide users on installation and usage. Finally, WAAS is licensed under Apache-2.0 and has a Code of Conduct and Security policy to ensure its users' safety.