Launching today
SKI
Free voice coding for Claude Code, Codex and more
353 followers
Free voice coding for Claude Code, Codex and more
353 followers
We just built SKI — voice coding for Claude Code, Codex and more. It's not dictation: your agent answers you out loud, like a real teammate, so you build at the speed you think. You can even bring it into a meeting to build live, or send it in your place to speak for you. It's an ambient thing that just sits on your desktop — hit a key, talk, it works. All on your machine, free. Available on Mac & Windows.
Talking is fasting than writing. But reading is faster than hearing (at least for me). How does the answering part work? Is it just a status update ("task finished") or does it say the agent's output out loud? We all know that agents can be very verbose in their answers.
SKI
@mateuszkonik Thank you for taking time to look into our product.
You can ask the agent to give a status update or a brief. The default instruction is to limit the talking to 2-3 sentences. But it is customizable and the agent gets to decide it based on user instructions. You could ask it to just share a status update. It will just update on the status on that case. You could ask it to give a brief or open a window and show the issue as well.
It doesn't read out the entire output. This is like a skill which the agent gets to use. It can decide what to tell the user. It can write 100 lines of code, and tell user that it needs your input to work on another task, out loud. So, 100 lines silent, and the request for help is spoken.
ClinicFrame
Will it run on agents running on isolated PCs without internet connection ?
SKI
@clemente_lopez1 Yes, ski runs completely offline once it is installed. No internet connection is needed for it to run.
Swytchcode
This is interesting. How do you highlight a specific code in a file using dictation? Just curious :)
SKI
@chilarai Multiple options
1. You could just say file name and the lines and LLMs now are intelligent enough to understand it.
2. You could open up the screen and just ask it to take a look at your screen - it can take screenshots and understand the lines. Or you could send screens using hotkeys along with your voice.
3. You could also talk just about the functionality implemented in the code and LLMs can understand them easily.
LLMs are now intelligent enough to understand and correct mistakes in transcription as well.
thats interesting, using voice to talk to ai and code is an amazng idea but do people actually want to talk or just type???
SKI
@aziz_mohd1 Maybe both. Speaking is much faster. And it is much easy to articulate things and explain things a lot in detail with speech.
For me personally, I just go away into doing other things when agent is being assigned a ask. For instance, I asked it to build something and it started doing it. But it takes a few minutes for it to complete. I don't feel good staring at the screen. So, I end up switching projects, or checking out phone. If I switch projects, I end up forgetting to check back. Maybe the agent is expecting me to clarify something and is waiting for my inputs.
Now, I connect multiple projects to ski and it just tells me if I need to check it out. It also keeps me updated on the progress. When I want to type, I can. I would recommend that you try this setup as well to make it more productive.
Something I would also recommend is the meeting connection. It can record meeting and connect it to your project, more like a meeting brain that you have.
Is the offline meeting recorder compatible with all meeting platforms?
SKI
@basil_lenin Thank you for checking out our product. In the offline meeting recorder, it records the system audio and your mic inputs. So, it will work for all meeting platforms or even websites where it plays a audio and you reply back. It runs on your device and is completely free.
Does SKI need API keys to use ?
SKI
@drew_nich Voice coding in SKI is completely free and doesn't need API keys from any voice providers for transcription or speech generation. Everything is local and on-device.
Does it diarize speakers in meeting recording ?
SKI
@anandakrishnan_vinod Thank you for checking our product.
There are 2 meeting recording options available
Offline meeting recorder - In this, there is no speaker diarization. It records and transcribes the audio from meeting and from your mic as separate channels. It can distinguish between what you said and what the meeting said, but cannot name the speaker from meeting. This feature is free and unlimited. Completely on device.
Agentic notetaker - This is a cloud feature. In this, the transcription has speaker names (as shown in the meeting). Supported platforms are - Teams, Meet and Zoom. The transcripts are stored into the connected project folder, which can then be accessed by the agent. Cloud features are billed per minute, with free 6 hours included. Needs sign up.