Just got benchmarks done for German langauge eval index i help maintain. Hy4 is a big improvement over h3 but still below Deepseek pro and Significantly below GLM 5.3 Flash .

hy4 ranks ~14th overall

https://dach.peerbench.ai/compare?models=tencent%2Fhy4-previ...

https://dach.peerbench.ai/models/tencent/hy4-preview

love the direction have asked myself manytimes why this feature sent doesn't exist somewhere; for sure would need to see some high profile people start using this to think its bullet proof

Maybe it's my dyslexic brain but "it's own frontier model" in my head converted to foundational model that was trained from scratch.

But this is qwen based.

But w/e I'm pro AI so more companies having more people with skills for more post training is cool

I tried to build a few non crypto decentralized apps. The killer was reliable always deliverability inside the browser. https://inbrowser.link/ was a huge jump in utility but came to late the ipfs.js just never worked consistently. In the end the only way to get good XP for yours users was that you provider the IPFS to http gateway for all content but then whats the point its just decentralization theater.

I remember in 2015 having the IPFS concept blow my mind its such a memorable moment when it really felt like someone designed something significantly different that current mainstream paradigms. But in the end it seems like it was still a case of a cool technology looking for a use-case not solving a real problem.

> In the end the only way to get good XP for yours users was that you provider the IPFS to http gateway for all content but then whats the point its just decentralization theater.

I gave my static photo album website image files paths that consisted of /ipfs/ as directory name and ipfs hash as file name.

Brave browser was able to figure out that this meant they were available on IPFS, and asked me if i wanted to load the files from IPFS instead of from the web server itself.

So over time if hosting files on IPFS got popular, it could have been used automatically by browsers later to load the files from IPFS instead of the web server of the site. Saving bandwidth for the website operators and distributing storage of content which might also help it remain available for longer time.

That sort of thing would have been one great outcome, for example, where this sort of hosting your own http gateway for your own IPFS files makes sense like I was doing. Mind you, the web server setup for this was completely straightforward and requires no running of the IPFS services on the web server itself. On the web server it really was plain old directories and file names. As long as they are real IPFS hashes. And calculating the IPFS hash itself I did on my laptop when adding photos.

Very sad to hear in this thread that Brave browser has actually removed IPFS support :(

I wonder if (in theory) it wouldn't make more sense to address this sort of thing at the system resolver level rather than in the browser, the way tor does.

Deeper OS integration would be nice, but harder to roll out. I think the plan early on was to make something useful/generic/agnostic enough that a couple of big browsers would add support, then it could try becoming a Web standard.

Unfortunately that hasn't panned out; but it still feels more feasible than getting Microsoft and Apple to add it to their OS network stacks.

Pretty much impossible because of the incentive structures OSes have. Microsoft and Apple want to surveil you and Linux is a fragmented bazaar.

Agreed. I was deep in the ecosystem, did the confs, shipped some apps. It’s hard to think of a use case that would be reliable now. Maybe some kinda file sync that is torrent like but http

also still getting used in some libgen sci-hub mirrors

oh man and when Brave removed native IPFS support. https://github.com/brave/brave-browser/issues/37735

Oh wow they did? I had no idea. Unfortunate

can this not be achieved through a system prompt adjustment ? Running yet another LLM call on top might not be the simplest solution.

Vampire survivor is an amazing game. If you can't accept the value of entertainment then the rest of your life is probably to easy.

But fair enough everyone can have their take.

We maintain German Langauge index as no one publishes or reruns these sepeartly.

Qwen 3.8 27B is a small improvement with some regressions in our benchmarks not a huge jump like benchmarks listed.

https://dach.peerbench.ai/compare?models=qwen%2Fqwen3.8-27b,...

German language has never been a big focus for asian models but they still outperform Gemma models https://dach.peerbench.ai/compare?models=openai%2FQwen%2FQwe...

So in production we have been using Gemini Flash Lite as primary and fall back to Qwen when gemini servers are overloaded or just giving us 429

Did you already test TranslateGemma? I use this model for my Android Studio Translation Plugin (https://plugins.jetbrains.com/plugin/30265-localizepipe) and so far it produces great results for its size.

If there are other models (of similar size) out there, that are better at this, please let me know.

From your benchmark, Qwen3.8 is nearer than Opus 4.8 than Qwen3.6. 0.1pp but still.

Also, a lot of people don't really care about german language capacity, maybe people programming in DDP idk.

PS: You benchmark seems saturated. Most values sit @>75% in a benchmark generally indicate that it's no longer as useful as a <70% one. I mean, Qwen3.8 is 77.5% and Fable5 80%, the poll of values is from 65% to 90%.

Same question.

I usually have one local clause orchestrating multiple remote Claude in different tmux . And then another orchester and remote vm worker sin tmux for another repo etc...

It gets a bit hard to keep the overview but I don't want to give up my parallelism, your too might help

Interesting, how is this orchestration set up?

One way is to use something like https://shellbox.dev - you start new boxes and manage it all through ssh, icl agent running inside