Hacker Newsnew | past | comments | ask | show | jobs | submit | verdverm's commentslogin

Buy access to the open models from a single US vendor like https://fireworks.ai

One company, multiple models, Fireworks is the fasts at making the models available (had GLM-5.2 before the other three we are evaluating)


Second this notion. After picking up an OEM Spark and running qwen36moe/dense, I was thoroughly impressed with what such small models can do and the (reasonable) speeds you can get. I'm back to using open weight models via an API (wanted more capability for the time being), but will be getting more hardware soon (re: ds4-flash and the fable shot heard round the world)

I'm using SearXNG, EXA, Tavily, and soon (tm) Cloudflare

They all give slightly different results, you can dedup / fusion with heuristics / another agent


There is a lower bar (that gets lower over time), but ime, the config you are describing is too low still.

qwen/gemma in the 27/35B range @fp8 are better than gemini-2.5, but less than gemini-3.1, you can run DS4-flash @fp8 on two DGX spark, and things keep becoming better. DiffusionGemma came out recently with 4x token gen speeds.

tl;dr - the models you appear to be trying with are too small or too quant'd


slow down and put more effort into vetting, trust your gut, make them share all screens during (coding) interviews

They explicitly do not want them. They will pay companies to abandon renewable energy projects that have already planned / started.

Not just climate data, they are changing how census data is aggregated so they can reverse-individualize for political goals, discussed 3 days ago.

https://news.ycombinator.com/item?id=48517377

https://www.npr.org/2026/06/12/nx-s1-5855734/census-bureau-d...


51° FOV, not much improvement here, but at least it is a much smaller form factor

Second this

related: Electrostates, Petrostates, and the New Cold War

https://www.youtube.com/watch?v=gLnxzkiB-GI


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: