Quick Demo Video (50s): https://www.youtube.com/watch?v=HM_IQuuuPX8
The goal is to get closer to natural conversation speed. It uses audio chunk streaming over WebSockets, RealtimeSTT (based on Whisper), and RealtimeTTS (supporting engines like Coqui XTTSv2/Kokoro) to achieve around 500ms response latency, even when running larger local models like a 24B Mistral fine-tune via Ollama.
Key aspects: Designed for local LLMs (Ollama primarily, OpenAI connector included). Interruptible conversation. Smart turn detection to avoid cutting the user off mid-thought. Dockerized setup available for easier dependency management.
It requires a decent CUDA-enabled GPU for good performance due to the STT/TTS models.
Would love to hear your feedback on the approach, performance, potential optimizations, or any features you think are essential for a good local voice AI experience.
The code is here: https://github.com/KoljaB/RealtimeVoiceChat
2025-05-05 20:53:15,808] [WARNING] [real_accelerator.py:194:get_accelerator] Setting accelerator to CPU. If you have GPU or other accelerator, we were unable to detect it.
Error loading model for checkpoint ./models/Lasinya: This op had not been implemented on CPU backend.
We are working hard toward upgrading the Wikipedia ZIMs, but it is far from being an easy feat. I'm mostly solo on this, and far from dedicating 100% of my time to this, so it does not move very fast. We are quite close to being able to reach the goal however, probably only a matter of weeks now.
Bonus: the tool will now get pretty good at making a ZIM of any Mediawiki, not only Wikimedia ones, we expect for instance to work on all Fandom wikis somewhere this year since there is significant knowledge over there.
The whole enchilada: https://download.kiwix.org/zim/wikipedia/wikipedia_en_all_ma...
Other versions: https://library.kiwix.org/#lang=eng&category=wikipedia
https://web.archive.org/web/20210729190016/https://support.s...
The original link is dead for some reason: https://support.startpage.com/index.php?/Knowledgebase/Artic...
TL;DR: Startpage appears owned by an ad company? https://web.archive.org/web/https://www.bizjournals.com/losa...
Could someone explain to me how an ad company and a privacy company work together? Seems like opposing interests?
Maybe Ecosia will be a good alternative later on: https://blog.ecosia.org/eusp/
Another suggestion would be https://searx.space/
>Because everything in Signal is end-to-end encrypted, we can rent server infrastructure from a variety of providers like Amazon AWS, Google Compute Engine, Microsoft Azure, and others while ensuring that your messages and calls remain private and secure.
First time I am seeing an organization against this. Kudos to them for standing up.
As good as Signal is I mean, you will want something under your control.