Spokenly
AI-powered dictation for Mac, Windows, Linux, and iPhone that lets you speak at the cursor using local models or your own API keys.
Short description
Spokenly is a cross-platform dictation tool that supports local speech models, live streaming, AI instructions, meeting recording, searchable history, and integrations with coding tools and AI assistants. Users can use local models for free or bring their own API keys.
Features
100+ languages: Includes mid-sentence mixing and automatic language detection
Local-only mode: 0 network requests
AI Instructions: Grammar fixes, filler removal, and app-specific formatting
Live streaming: Words appear as you speak
History: Searchable and exportable history with audio replay
Meeting recording: Captures microphone and system audio without a bot and transcribes later
Pricing
Local Models: $0 forever, with Whisper and Parakeet, offline use, unlimited dictation, and no account
Your API Keys: $0 forever, including local features plus OpenAI, Deepgram, and Groq real-time models
Pro: $8.33/mo, billed at $99.99/year, with cloud models, sync, and AI cleanup
Refund: 14-day refund
Integrations
MCP: MCP server for agents
Voice coding: Claude Code, Codex, and Cursor
AI assistants: ChatGPT, Claude, Gemini, and Perplexity
BYO providers: OpenAI, Deepgram, and Groq
Platforms
- macOS
- Windows
- Linux
- iOS
What users love about Spokenly
Free local models: Unlimited use without an account
Offline speed: Users describe offline models as much faster than Apple's built-in dictation
100+ languages: Automatic language detection
What users complain about Spokenly
Cloud models: Require Pro or BYO keys because the free tier is local-only
Meeting functionality: No true meeting notes or summaries, only recording
Mobile coverage: iOS is newer and Android is missing
