Hacker GitHub

← All repos

Niko1221/Strata

View on GitHub ↗ ★ 38,071 C++

Qwen3.8-Flash-Next on any consumer hardware: one-click install for Windows / Linux. Strata inference engine, OpenAI/Anthropic API on localhost, optional image input.

19 mentions on Hacker News by 17 people: 3 stories, 16 comments · first 2026-09-28 · latest 2026-10-31

Mentions per year 19 mentions · busiest year 2026 (19)
20082026
AllStoriesComments NewestTopOldest
  1. Yep, I agree. You can already run them locally on a top-shelf gamer PC with decent token throughput: https://github.com/Niko1221/Strata That's today. As for tomorrow, consider that literally every chip maker is working on better and better…
    comment by skrellm · 2026-10-31 · in A tiny world model running games on a Raspberry Pi Pico 2 W
  2. …most of them still had a >30% chance of producing wrong/non-working replies. I then tested Strata ( https://github.com/niko1221/strata ) and the results are not even close. You can basically ignore everything in that table and use Strata…
    comment by plqbfbv · 2026-10-27 · in New iMac with M6
  3. …(64GB), you can run Qwen 3.8 Flash Next with a GPU as small as 12GB, this repo is pretty amazing: https://github.com/Niko1221/Strata/ My setup is a Ryzen 3600, 90GB DDR4 allocated to a VM (don’t even really need that much, just had it…
    comment by Grombobulous · 2026-10-27 · in New iMac with M6
  4. Take a look at the docs for https://github.com/niko1221/strata for some ideas, particularly https://github.com/Niko1221/Strata/blob/main/docs/MODELS.md#... .
    comment by CamperBob2 · 2026-10-26 · in OpenAI and Anthropic cut their LLM prices by up to 80% within a few months. Why?
  5. …if you have 64+ gb ram and use strata or a strata fork (vram not super important with this method). https://github.com/Niko1221/Strata
    comment by eyegor · 2026-10-20 · in MiniMax M3.1 Flash
  6. …ram 16gb AMD 7900XTX. It goes about 20tok/s off the ssd. I'm guessing 64gb would go to 40tok/s. https://github.com/Niko1221/Strata Quality wise its not fast enough for me to eval. I'll probably spend some tokens on it through openrouter.…
    comment by drwiggly · 2026-10-19 · in DeepSeek V4.1 Flash 1017 (Max)
  7. You can get something relatively capable and usable with your configuration and Strata: https://github.com/Niko1221/Strata
    comment by lhl · 2026-10-19 · in DeepSeek V4.1 Flash 1017 (Max)
  8. A Strata ( https://github.com/Niko1221/Strata ) challenger! The first one I've seen (although I'm sure there are and will be more as more time progresses...) It seems to require a bit more system memory (70GiB free RAM as opposed to 32GB)…
    comment by peter_d_sherman · 2026-10-17 · in Show HN: Capy-flash – a 125B MoE model on two 16 GB consumer GPUs
  9. …the active count + a few experts (although it will run faster the more experts you can fit). [0] https://github.com/Niko1221/Strata
    comment by eyegor · 2026-10-15 · in LFM2.5-24B-A2B: Built for fast on-device agents
  10. Then check out those: https://github.com/prabanta-dev/janas <- this, focuses on Qwen3-Next-80B-A3B. https://github.com/Niko1221/Strata <- Qwen3.8-Flash-Next. I use this (tuned) and I cannot understate how satisfied I am.…
    comment by MadameMinty · 2026-10-13 · in Janas: Run LLMs larger than RAM by streaming experts from SSD
  11. …(Jumanji) What year is it? What a waste of time. Strata already does this with Qwen3.8-Flash-Next: https://github.com/Niko1221/Strata
    comment by suprjami · 2026-10-13 · in Janas: Run LLMs larger than RAM by streaming experts from SSD
  12. …to run the 180B Owen model on strix halo using strata and even get reasonable speed out of it. https://github.com/Niko1221/Strata
    comment by throwa356262 · 2026-10-09 · in Step 5 Preview, a 1M-context MoE from StepFun, shows up on OpenRouter
  13. what they should freak out about is qwen 3.8 github.com/Niko1221/Strata it rips with just 64 ram and a 9070xt
    comment by nicman23 · 2026-10-09 · in Why isn't the industry freaking out about DeepSeek 4.1 Flash?
  14. > Set up Strata on this PC for me: https://github.com/Niko1221/Strata - follow docs/AI_SETUP.md in that repository. And I thought piping to bash was bad
    comment by deadbunny · 2026-10-04 · in Run Qwen 3.8 Flash Next (125B) on consumer hardware (RTX 4090) at 100T/s
  15. They have some community benchmarks published https://github.com/Niko1221/Strata/tree/main/bench/results
    comment by snehesht · 2026-10-04 · in Run Qwen 3.8 Flash Next (125B) on consumer hardware (RTX 4090) at 100T/s
  16. …91% of the full model's SWE-bench Verified score (measured by its authors) and fits 32 GB of RAM. https://github.com/Niko1221/Strata#which-model-should-i-pick
    comment by mkl · 2026-10-04 · in Run Qwen 3.8 Flash Next (125B) on consumer hardware (RTX 4090) at 100T/s