# Whisper API Server

Whisper API Server is a self-hosted drop-in replacement for OpenAI's Whisper speech-to-text Application Programming Interface (API), giving homelab users a local OpenAI-compatible endpoint for audio transcription. It is open source and self-hosted, installable on Docker Hub and Unraid.

Canonical page: https://ourlocalhost.com/app/2247/whisper-api-server

## Description

Whisper API Server is a self-hosted drop-in replacement for OpenAI's Whisper speech-to-text Application Programming Interface (API), giving homelab users a local OpenAI-compatible endpoint for audio transcription without sending audio to external services.

## Platform availability

Available on 1 of 10 platforms: Unraid.

## Health score

43/100 (Fair), based on 4 of 5 factors.
- Maintained: 72 (last commit 13 days ago)
- Popular: 25 (17 GitHub stars)
- Easy to install: 35 (1 install platform, install notes, Docker image)
- Security check: 15 (2 serious findings, 6 minor, CI, tests)

## Security check

2 serious findings: Uses anyio 4.13.0, which has a known critical vulnerability; Uses certifi 2026.2.25, which has a known critical vulnerability.
Automated checks on commit fcb2e66, 18 Nov 2026. Not an audit.

## Links

- GitHub: https://github.com/didmar/whisper-api-server
- Docker Hub: https://hub.docker.com/r/didmar/whisper-api-server
- Unraid: https://ca.unraid.net/apps/whisper-api-server-05rkzv70mqtknb?q=Whisper-API-Server

## Common questions

### How do I run Whisper-API-Server with Docker Compose?

Use the command docker compose up --build -d to build and start the service.

### How do I increase concurrency in Whisper-API-Server?

Adjust the workers option when running the container. Use more than one worker only if you need to handle multiple requests in parallel.

### How do I keep container cache data across restarts?

Mount a local .cache directory to /root/.cache in the container with read-write access. This is the cache path shown in the Docker run example.

### Where can I see the Whisper-API-Server API reference?

The server exposes API documentation at the /docs path on port 8000.

### What endpoint should an OpenAI-style client call for transcription?

Post a multipart form to /v1/audio/transcriptions. Include the audio file as file and set model to whisper-1.

## Alternatives

- FluidVoice
- Handy STT: Free, open source speech-to-text app for desktop, works fully offline, prioritizes privacy, extensibility, cross-platform support, and acce…
- Speech Note: Note taking and reading with Speech to Text and Text to Speech.
- Spokenly: Real-time voice dictation for Mac with instant transcription, customizable shortcuts, advanced voice recognition, no data collection, and b…
- TranscribeX: macOS app for local AI-powered transcription, translation in 100+ languages, speaker labeling, and GPU acceleration, with no cloud or data…
- Txtify: Txtify is an open-source, containerized web application that transcribes and translates audio from YouTube videos or uploaded media files,…
- Vibe Transcribe: Provides fully offline audio and video transcription with local AI models, batch processing, real-time preview, multi-format export, transl…
- Voxtral: State-of-the-art speech models with transcription, translation, and audio understanding, available via API or self-hosted, optimized for co…
- scriberr: Scriberr is an open-source, self-hosted audio transcription and summarization tool that runs OpenAI Whisper models locally to convert audio…
- speakr: Speakr is a self-hosted AI transcription and note-taking platform for individuals and privacy-conscious groups that turns audio recordings…
- whishper: Transcribe any audio to text, translate and edit subtitles completely locally with a web UI.

---
Source: Our Local Host (https://ourlocalhost.com/app/2247/whisper-api-server), the self-hosted and open-source app directory.
