ndom91

Sorry opencomputer :sweat_smile:

But made a local fork that works just as well with 'claude-code -p' and a local copy of ffmpeg

https://github.com/ndom91/shipvideo-local

Are you in the US? Hehe

You don't say?! :joy:

This applies to pretty much everyone who isn't the US/China, no? I guess Russia also has a good amount of home grown web services (Yandex, Wildberries, etc), but not much in the AI space that I'm aware of.

Been wanting to try this out on top of https://github.com/TristanBrotherton/voicepe-realtime with the HomeAssistant VoicePE

Anyone else hack something together with HA / the Voice PE yet?

Looks like it needs a second model to do function calling, which gpt-realtime-2.5 didn't, and the Voice PE XMOS chip's audio pipeline might not be a great fit for full duplex back and forth, like what gpt-live-1 now supports.

Is archive.is/ph having issues for anyone else lately? Seems they've run out of captcha credits and i can't get past that initial captcha for the past few days :/

Upgrade from 10.11.11 was also successful here - only took 2-3 minutes. Can definitely notice the performance improvement

Wait what haha. Do you also own samiam.org?

What's the disk space usage like? i.e. average per-day/week additional storage in your usage?

Out of curiosity, what's the use case for ringconnectd? Just to find out what your IP is? What would you need that for back in the day?

I'm a bit of a privacy nerd, but even so it still took me a while to convince myself it was worth paying for a search engine.

Ever since getting over that, however, I've been extremely happy with Kagi. Couldn't recommend it more.

Using qwen 3.6 27b for local coding as well and downloaded Laguna s 2.1 but haven't had time to give it a full spin yet.

Curious for any more experiences

"AI is progressing at a rate that our society might not be ready for. Frontier labs are very close to AI that can exceed even the best people on almost every metric of intelligence. This will lead to unprecedented social and safety risks. To ensure a positive future we need to develop AI in a way that is driven by responsibility and thoughtfulness."

Responsibility and thoughtfulness? In this economy?

The sad part with all these social media age restriction laws is that they'll by definition require some sort of age verification mechanism that will require us all to prove our age (15+ in the case of France).

I wish the world would have gone down the route of banning social media in schools instead. But I guess that wouldn't create an execuse to de-anonymize the web :(

Again not released on huggingface immediately?

Seems like it would be super useful for folks new to development and version control to have a visual representation of what's happening with your commits

Congrats! Better auth still provides a grade A dev experience, even with all the plugins, integrations, and tons of things they support.

Best of luck over there!

Wow the prices on these have really come up.. Got my Framework desktop mainboard (Just the motherboard + CPU + soldered 128gb RAM) in Dec 2025 for ~1900 EUR

Been working on this side-project for the last few months and finally put together a nice marketing page and docs site.

It's a beautiful dashboard and proxy for monitoring models, requests, and API keys. It also supports routing rules, and ,proxy metrics, playground tabs, additional attribution headers, and much more. In addition to local models, it also supports proxying anthropic / openai requests from local coding agents like claude-code so you get much more insight into what they're doing too.

Currently relies on llama-swap / llama.cpp as the underlying inference engine, but that part is designed to be very generic and easy to swap out / build support for additional inference applications like ollama, vllm, etc. as long as they expose similar APIs.

It's entirely self-hostable without any cloud services. I'd love to hear any feedback you all may have!

+1 using llama.cpp Vulkan releases with the Qwen models - runs much better than the ROCm releases.

I'll have to give the preserve_thinking a shot.

Not 100%, I still fall back to Claude for most day-job stuff. But I've been trying to use Qwen 3.6 and Gemma 4 on my framework desktop mainboard (Strix Halo) as much as possible.

I've been working on an ops style tool for local LLM inference. Proxying, api keys, request logging, model rewriting and much much more.

https://github.com/ndom91/llama-dash

Long time protonmail subscriber and reasonably happy, but them not allowing SMTP access (other than through proton bridge which is a GUI app - workaround: https://ndo.dev/blog/headless_protonbridge) kind of sucks, and their search functionality is also not great.

Debating moving over to Fastmail as well

Are they really not doing huggingface releases anymore? I remember not being able to find one on their latest HN front-page release either :(

Another happy user here - it's the power users chromium fork. Criminally underrated. It's a small Norwegian team with no VC funding and a sustainable business model.

I understand if you want to stick with Firefox, but until Ladybird and co are ready for prime time, I'm sticking with Vivaldi.

This major release bump is a bit disapointing though. Was expecting some more headlining features than just a bit of a UI clean up.

Is this one of those ones where they'll drop the huggingface release a week later? Or do we know for sure that this is staying proprietary?

How does your "Crisper Speech Model" compared to parakeet-0.8b or similar? I've had great experience with that model via handy.computer.

Claude. Check out the recent "the unusual effectiveness of HTML" articles on HN recently.

So happy to see this climbing HackerNews. I hope someone at Bambu gets their head out of their ass after this debacle, but I'm not holding my breath

Where's the gguf from Unsloth and co?

Anyone know much about the new top of the line Intel vs AMD CPU options? Which is more power efficient? Powerful?

`/model claude-opus-4-6`