AudioBud

AudioBud is a dictation app for Windows that puts speech into the text field you're writing in. It builds on Handy by CJ Pais and suits people who want voice typing with local speech recognition rather than a hosted transcription service.

• Dictation across apps: Use toggle or hold-to-talk recording and insert the transcript into the focused app through clipboard paste or direct typing.
• Local speech models: Choose Parakeet, Whisper, Moonshine, SenseVoice, GigaAM, Canary or Cohere, or use custom Whisper GGML files; options include English-focused recognition, East Asian languages, Russian and multilingual translation.
• Vocabulary and formatting: Add custom words and replacements for names or recurring mishears, format spoken numbers, or keep raw transcripts.
• History and personalization: Retry or save transcriptions and control recording retention, with optional on-device vocabulary learning from your dictation history that you can review, export or reset.

Speech recognition works offline, keeps audio on your machine and doesn't require an account. Parakeet V3 is the Windows default and supports DirectML acceleration without requiring CUDA or an NVIDIA GPU.

Optional text cleanup uses a separate LLM service, such as OpenAI or Anthropic, or a custom OpenAI-compatible endpoint. Choosing a cloud provider sends text to that service for processing rather than keeping that step within the local speech model. AudioBud is open source under the MIT license and is available for Windows x64 through direct downloads and the U.S. Microsoft Store.