Launching today
MachinesFluent
AI powered dictation with extreme flexibility
5 followers
AI powered dictation with extreme flexibility
5 followers
AI smart dictation for power users focused on extreme flexibility. Windows desktop first. Local (no Wi-Fi) or cloud. Hundreds of models via BYOK or direct Sign-in (Oauth). Combine Transcription models, Language models, and custom prompts for insanely accurate, context-aware outputs in 100+ languages. Process and transform clipboard content, including images. Search the web by voice. Summarize long text and get rich responses, including diagrams.
Hi Everyone.
Long story short, my atrocious chronic back pain prevents me from sitting straight and typing for extended periods of time, so I solved the problem by building a dictation tool that works anywhere your keyboard does. The app eventually became a bit of an AI powered Swiss army knife, as I extended the features.
Speaking is not just faster, it changes the way you work. Raw transcription will likely always produce occasional errors, but a second step in the pipeline using AI (LLMs) for reformatting, fixes the issue.
If i had to define this app with one word ---> Flexibility.
- You prefer data privacy? Use offline "local" models (Ollama & LM Studio supported).
- You prefer the convenience of cloud services? Use cloud models.
- You want a specific writing style when using a specific app or website? Create your own mode.
- You prefer AI "company X" over "company Y"? We support them all and instantly access their latest models.
- You work across various domains with their own jargon? Create multiple dictionaries.
- You seek powerful workflow customization? Combine prompts with composable context artifacts.
- You don't like holding a button several minutes when you have a lot to say (I hate it)? Use streaming mode.
- etc, etc.
There is no one size fits all. Everybody works differently. Make the tool adapt to you, not the other way around.
Flexibility doesn't mean complexity. I tried my best to keep it simple while preserving power features.
Notable Features:
- Adapts the output style to the app you're using (e.g casual for Whatsapp, professional for Slack, etc)
- Anything you copy in the clipboard (text & image) can be used by AI to reformat, extract, explain, expand, etc.
- Ask anything and get an answer immediately instead of switching tabs or apps (Web search included).
- Decouples transcription from AI reformatting, to save you from repeating yourself.
MachinesFluent market positioning is essentially:
- We don't compete with top-tier models, we aggregate them instead and keep the list updated.
- We don't care about your data. No account registration needed.
- High degree of customizability for power users.
- Lifetime offer over monthly fees (our monthly offer mainly exists for you to try). Paying monthly to simply talk to your computer seemed unnecessary.
About the other solutions out there (I've tried almost all of them):
- Open source solutions are free but rudimentary and feel like weekend projects.
- Proprietary transcription models are cloud-based so you rent them but can never own them by paying once.
- The vast majority lack sufficient customizability for power users.
- The alternatives are worse if you're on Windows.
I genuinely hope you'll find this app useful and sufficient for your dictation needs.
Thank you to everyone who will take the time to try it.
NB: Windows first support for now. MacOS considered as next OS support. No ETA.