Spokenly, developed by Vadim Akhmerov, is an AI voice input (dictation) app for macOS and iOS. It transcribes your speech into text in real time and types it directly into any text field — email, chat, editors, and more. Alongside cloud models such as OpenAI and Deepgram, it supports local models including the Apple speech analyzer, Whisper, and Parakeet; with local models it runs entirely offline and free of charge. It also offers automatic detection of 100+ languages and AI grammar correction. The official site also lists Windows and Linux versions.
Key Features
- System-wide dictation in any app: Press a shortcut key and speak — the transcription is typed straight into the focused text field. No app switching or copy-pasting required
- Free, offline operation with local models: Local models such as Whisper and Parakeet run offline on Apple Silicon and can be used free of charge without limits. Since audio never leaves your device, they are also reassuring privacy-wise
- Cloud models with your own API keys: If you bring your own API keys for OpenAI, Deepgram, Groq, Anthropic, Google, and others, you can use high-accuracy cloud models while staying on the free plan (BYOK)
- Automatic detection of 100+ languages: Spoken language is detected automatically. Even when Japanese and English are mixed, no manual switching is needed
- AI text processing and grammar correction: Transcripts can be polished with AI grammar fixes and custom instructions, turning spoken words into well-formed text
- Integration for AI coding: It ships with an MCP server, supporting integration with AI coding agents and voice-controlled Mac automation (Agentic Actions)
Pricing
| Plan | Price | Highlights |
|---|---|---|
| Free | $0 | Unlimited transcription with local models, cloud models with your own API keys |
| Pro | $9.99/month | Managed cloud models with no API keys required, priority support. One subscription covers both Mac and iPhone |
Pricing is as of August 2026. Please check the official site for the latest pricing.
Pros & Cons
✅ Pros
- Even on the free plan, local models allow unlimited transcription, so you can use it daily at zero cost
- Local models never send audio off your device, making it usable for confidential work
- Bring-your-own-key (BYOK) support lets you flexibly choose your preferred cloud models
- Automatic detection of 100+ languages handles multilingual input
- A single Pro subscription covers both Mac and iPhone
⚠️ Cons
- Comfortable local-model performance assumes an Apple Silicon Mac
- Using cloud models without API keys requires the Pro plan ($9.99/month), which is not the cheapest among similar apps
- With your own API keys, each provider’s usage-based API fees apply separately, and setup requires some knowledge
- As with dictation apps in general, recognition accuracy for technical terms and proper nouns varies depending on the model
Comparison with Similar Services
| Item | Spokenly | superwhisper | Wispr Flow | Aqua Voice |
|---|---|---|---|---|
| Platforms | macOS / iOS (Windows and Linux also listed) | macOS / iOS | macOS / Windows / iOS | macOS / Windows |
| Local models | Supported (free, unlimited) | Supported | Not supported (cloud-based) | Not supported (cloud-based) |
| Bring your own API key (BYOK) | Supported | Not supported | Not supported | Not supported |
| Free plan | Yes (unlimited local) | Yes (limited) | Yes (weekly word limit) | Yes (limited) |
| Paid plan | $9.99/month | Paid plans available | From $12/month | Paid plans available |
Who Is It For
- People who speak faster than they type and want to draft emails, documents, and chats by voice
- People handling confidential information who want local processing that never sends audio off the device
- People who want to start using dictation daily for free (local models are unlimited at zero cost)
- People who already have OpenAI or Deepgram API keys and want high-accuracy transcription with their preferred models
- Developers who want to dictate instructions to AI coding agents without breaking their flow
Summary
Spokenly is a dictation app built around free, unlimited transcription with local models, while also offering bring-your-own-key cloud access and managed cloud models. Its strength is flexibility: go local for privacy, or cloud for accuracy, depending on how you work. Start by replacing your everyday typing with the free plan’s local models, and consider the Pro plan or BYOK if you want higher accuracy or less setup.