By Denys MedvedievSetup Whisper Cloud Mode
Use OpenAI for transcription and AI enhancement. High accuracy, web search, GPT-5 text processing — all powered by your OpenAI API key.
Last updated: April 2026 · Applies to v2.1+
Get an OpenAI API key
Cloud mode requires a personal OpenAI API key. Sign up at platform.openai.com, click "Create new secret key", copy it (starts with sk-).
You pay OpenAI per use, not Whisper. Typical cost: $1–$3/month for moderate dictation.
Switching to cloud because stock Mac Dictation broke? Read what to fix before paying for an API key — most of the time it’s a permission toggle, not a missing tool.
Install & set the key
Open Whisper, go to Settings → Model, switch to the Cloud tab, and paste your API key. The key is encrypted on your device.
sk-…xxxxxxxxxxxxxxxx
Pick a transcription model
Two options from OpenAI: GPT-4o Mini Transcribe ($0.003/min, fast) or GPT-4o Transcribe ($0.006/min, highest accuracy).
Enable AI enhancement
AI text improvement automatically removes filler words, fixes grammar, and polishes your dictation into publication-ready text.
Web search by voice
Toggle web search on. Ask a question — Whisper detects it needs live info, searches, synthesizes sources, and pastes a researched answer with citations.
Useful for fact-checking, research, getting up-to-date info — without leaving your editor.
Custom AI presets
Create reusable instruction presets: "Rewrite as professional email", "Summarize in 3 bullets", "Translate to Spanish". Pick one before recording — AI processes accordingly.
Start with cloud mode
Higher accuracy, web search, and AI text enhancement. Cloud features are part of Whisper Pro — your OpenAI API costs (typically $1–$3/month) are paid directly to OpenAI on top of the subscription.
