mkiol
Are you planning to add an API?
Hello, dear developer mkiol. Are you planning to add OpenAI Compatible API to your awesome application in the near future? If not, then I ask you to do so. In my opinion this would be incredibly useful. This would allow Speech Note to be used with an incomparably larger number of applications than it currently has. It seems to me that the API will soon become a standard for all similar applications. I continue to use your wonderful app and thank you every time I open it for choosing to create it. And you did it so well. Many thanks for your work, respect! I finally use the Russian language in Speech Note and I am completely satisfied with the quality of the translation of the Speech Note interface.
Hi, thanks for your question and thanks for the kind words about Speech Note :)
Are you planning to add OpenAI Compatible API
Could you elaborate a bit on this topic? What kind of API do you have in mind?
In general, Speech Note is a "fat" GUI app that is designed for direct use by the users. The primary use-case is not integration with other applications. If someone is looking for integration, a better way is to try to play with whisper.cpp or piper for example.
I meant an API that will allow artificial intelligence systems and other programs to use Speech Note and all its functions. For example, such an API as in KOBOLD CPP.
This KoboldCpp is an interesting project I must say. The API is HTTP REST.
Currently Speech Note has very limited external API exposed via DBus. You can use this API to call an action such as "start-listening" or "start-reading" (full description is in the Settings->Accessibility). Perhaps extending this API and adding an HTTP REST option could be useful. I will think about it :)
Do you have a specific API use case in mind?
Do you have a specific API use case in mind?
No, no, it was just a wish. This is just a suggestion. I would be comfortable using this to integrate with other applications. Including for AI
Some kind of API would be really nice. My use case would be to run Speech Note in background while gaming and use it for dictation that could than be used in game for chat with other players. My vision is that this could be really nice on handheld PCs (Steam deck for example). Instead of typing on touch keyboard, I would just like to open on screen keyboard interface, dictate and than output from Speech note could be again available for inputing to game. In my mind ideal would be to copy output to clipboard like when someone does Ctrl+C and directly paste it after dictation to game trough keyboard by simulating pressing of CTRL+V. This way it would be pretty independent on game itself as it would pretty much simulate user typing it or rather pasting it by himself.
This would greatly speed up and simplify interactions in multi-player, specially MMO type of games, as I said, specially on handheld where keyboard is just emergency type of tool.
Direct Use of whisper.cpp sounds good, but Your app is nice and usable for everyone, Anybody can choose model lang. etc, all comfortably from UI and there fore it makes it kind of more approachable for average user compared to alternatives.
I would just try to write app to watch for on screen keyboard, enable listening over api, wait for response in form of output text, copying this text to clipboard, than pasting it wherever trough keyboard itself. User would than just push enter to send pasted message. This would be my usage of API.
