aegis_camera

4 ms·

1. 31.▲
   by aegis_camera 7mo ago

   You do have a deep understanding of AI hardware landscape. Thanks for your analysis.
2. 32.▲
   by aegis_camera 7mo ago

   M5 MAX should be very capable, you have a great brand new MBP.
3. 33.▲
   by aegis_camera 7mo ago

   It's a subset task. ..
4. 34.▲
   by aegis_camera 7mo ago

   When you are not at home, you can send your message to your dashboard agent for your query. This is one use case I found.
5. 35.▲
   by aegis_camera 7mo ago

   You are right. I have Mac mini M2 16GB, it does hold all the cameras I have. Small models like QWEN 9B + LFM 450M handle their security job nicely with < $400 budge. Will extend the test to more model and thanks again for your insight.
6. 36.▲
   by aegis_camera 7mo ago

   :)
7. 37.▲
   by aegis_camera 7mo ago

   Thanks a lot for your feedback :) I've noticed the slow down of QWEN3.5, so I turned it off thinking mode, the thinking mode even count words like ( 1 count 2 the 3 words, lol which is very funny ). You are very correct, I just have 2
8. 38.▲
   by aegis_camera 7mo ago

   Thanks for your insight, hardware of AI will be cheaper and memory of footage would be always saved locally.
9. 39.▲
   by aegis_camera 7mo ago

   Thanks for pointing out Unifi Protect, as long as the camera supports ONVIF(RTSP), then it could be connected, please let me know more, I'm not familiar with Unifi Protect, will do more research...
10. 40.▲
    by aegis_camera 7mo ago

    Aegis is able to connect to ONVIF camera, save motion triggered clips. Apply VLM pipeline for context understanding. It also helps to download video clips from BLINK/RING cameras, so you have persistent memory of all your video clips l
11. 41.▲
    by aegis_camera 7mo ago

    This a good platform. I was thinking about to get one
12. 42.▲
    by aegis_camera 7mo ago

    I have also 4070 laptop version during heavy discount season, before 50series came. And upgrade to 96GB DDR5 when it's cheap... So I like LFM 450M + QWEN 9B Q4, they are good fit to 8GB VRAM.
13. 43.▲
    by aegis_camera 7mo ago

    Exactly, the memory of full context is very personal, so I'd like to keep the local.
14. 44.▲
    by aegis_camera 7mo ago

    LFM 450M for vision task, QWEN 9B Q4 for Orchestration, this provides a good result.
15. 45.▲
    by aegis_camera 7mo ago

    My prefer is LFM 450M for vision task, QWEN 9B Q4 for Orchestration
16. 46.▲
    by aegis_camera 7mo ago

    Entry level is actually MAC MINI 16GB at <$499, I have models running on M2 MINI 16GB, it's working with small models.
17. 47.▲
    by aegis_camera 7mo ago

    Will also test DGX SPARK which I have.
18. 48.▲
    by aegis_camera 7mo ago

    Memory is the limitation, M5 has larger memory options. So large language model could be used.
19. 49.▲
    by aegis_camera 7mo ago

    The M5 Pro just dropped, so here's a real AI workload instead of another Geekbench score. We run Qwen3.5 as the brain of a fully local home security system and benchmarked it against OpenAI cloud models on a custom 96-test suite. The Q
20. 50.▲
    MacBook M5 Pro and Qwen3.5 = Local AI Security System(sharpai.org)

    173 points by aegis_camera 7mo ago|152 comments
21. 51.▲
    Show HN: CameraClaw – record and review what OpenClaw did in its sandbox(github.com)

    3 points by aegis_camera 7mo ago|0 comments