I wish I had known about Pipecat a lot sooner. I found out about it a few weeks back, and since Gemma 4 launched, I've been building my own entirely local voice assistant using Gemma 4 + Kokoro TTS + Whisper from scratch - https://github.com/pncnmnp/strawberry.
Pipecat's smart turn model is really good for VAD - https://huggingface.co/pipecat-ai/smart-turn-v3
Yeah Gemma4 was and is great fun to do this with - I too am building pretty much the same as yourself in Go.
https://github.com/zarldev/zarl & https://www.zarl.dev/posts/hal-by-any-other-name
Looks like everyone is building one of these, I have my own little version that's using streaming STT, it can actually be too fast in some cases, and I have a little ring buffer grabbing audio from before the wake word detection fires (so it can hear "Hey Jarvis, turn on the lights" without deliberate pause) https://github.com/jaggederest/pronghorn/
What do you have going on the hardware side? I want to plug this into hass but don’t know what hardware I need for reasonable latency
The whole setup works on my M2 MacBook Pro with 16 GB RAM. I use Gemma 4B via LiteRT-LM.
I've found that LiteRT-LM has a much lower DRAM footprint than Ollama. I've also made tons of optimizations in the code - for eg, you can do quite a bit with a 16k context window for a voice assistant while managing a good footprint, so I keep track of the token usage and then perform an auto-compaction after a while. I use sub-agents and only do deep-think calls with them, so the context window is separated out. In a multi-turn conversation, if Gemma 4 directly processes audio input, the KV cache fills up within a few turns, so I channel it all via Whisper.
Also, by far the biggest optimization is: 3-stage producer-consumer architecture. The LiteRT-LM streams tokens and I split them into sentences. A synthesizer thread then converts each sentence to audio via Kokoro TTS - the main thread then plays audio chunks sequentially. There's a parallel barge-in monitor thread. https://github.com/pncnmnp/strawberry/blob/main/main.py#L446
I did not want to use openWakeWord or Picovoice because they had limitations on which wake word you could choose. Alternative was to train a model of my own. So I created my own wake word detection pipeline using Whisper Tiny - works surprisingly well: https://github.com/pncnmnp/strawberry/blob/main/main.py#L143...
Also, I have VAD going with smart turn v3 (like I mentioned above) + I use browser/websocket for AEC + Barge-in (https://github.com/pncnmnp/strawberry/blob/main/audio_ws.py).
I'm using the MacBook's built-in microphones for this, though, and I haven't fully tested it with other microphones. I've been ironing out the rough edges on a daily basis. I should write a quick blog on this too.
Check out [0]. You can do 'Voice AI' on small/cheap hardware. It's the most fun you can have in the space ATM :) It's been a while, but posted a demo here [1]
[0] https://github.com/pipecat-ai/pipecat-esp32
[1] https://www.youtube.com/watch?v=6f0sUEUuruw
beautiful demo - is it running fully locally or talking to 3rd party API’s? That box was jaw dropping small
For the best experience, you'll still want it to communicate with 3rd party APIs to handle the speech to text, text to speech, and LLM.