Welcome to TinyProgrammer Discussions!

[cuneytozseker]

If you've built a TinyProgrammer, I'd love to hear from you. Share your setup, suggest features, post photos of your enclosure, or just say hi.

A few starter questions:

• What Pi and display are you running?
• Which LLM model has been your favorite?
• What's the best program your device has written?
• Have you 3D printed the enclosure?

Hope to hear from you.

[pageauc]

I thought your project was interesting and decided to try it. Have not started yet. I have written some of my Github projects using DeepSeek https://github.com/pageauc . Tried other LLM's but found deepseek at the time was best for me. I have found that as an AI coding project progresses, the LLM sometimes forgets some previous detail or prompt thus causing the project to go into the weeds until you resolve the cause but otherwise it can produce some impressive code. Over the last year or so most of my efforts have been upgrading to python Picamera2 and compatibility with Bookworm/Trixie. Other than that I have implemented various docker projects like Calibre-web, Audiobookshelf, mstream, pi-gallery, rpi-monitor Etc A bit of work weeding out the inferior docker implementations. Gave openclaw a try and HomeAssistant Automation but on the back burner. I have quite a few projects that I can't be bothered putting on github but may get AI to help me document and refine them with install bash scripts for user curl install. I have several rpi 4's laying around and thought I would give your project a try. My choice would be to use docker and a small 1080p hdmi display stored away. I would like to see if I can implement using DeepSeek. FYI I have been retired for 20 years and just like to do this as a hobby. Regards Claude ...

…

On Sat, Apr 11, 2026 at 12:47 PM Aerovisual ***@***.***> wrote: If you've built a TinyProgrammer, I'd love to hear from you. Share your setup, suggest features, post photos of your enclosure, or just say hi. A few starter questions: - What Pi and display are you running? - Which LLM model has been your favorite? - What's the best program your device has written? - Have you 3D printed the enclosure? Hope to hear from you. — Reply to this email directly, view it on GitHub <#10>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABNPKZDXRWCKVSEOQD2I7C34VJZKBAVCNFSM6AAAAACXVCEDXWVHI2DSMVQWIX3LMV43ERDJONRXK43TNFXW4OZZHA3DINBVG4> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

-- YouTube Channel at https://www.youtube.com/user/pageaucp <http://www.youtube.com/user/pageaucp> GitHub Repository at https://github.com/pageauc

[cuneytozseker]

Thank you! DeepSeek V3.2 is already in the model rotation via OpenRouter so you're covered there. The Docker setup should work well with a 1080p display via the MJPEG stream.

[clunkclunk]

It's a rainy Saturday here so I thought I'd put one together for fun.

Is there any performance benefit to different Pi models? I have a few I can choose from: 3B w/1GB RAM, 4B w/1GB RAM, 4B w/4GB RAM, 5B w/1GB RAM. I have a 1080p portable monitor (hdmi) I'll use with it.

[cuneytozseker]

Currently I'm running this on rpi4 and a pizero2w so any of them should be enough for cloud api. If you want to experiment with local model then 4gb+ is needed I think.