Issues

• Which model next? Vote here

  VecSzn/PolyStrata#16 · Filter by author VecSzn opened on Oct 12, 2026
  6

• GLM 5.3 Flash vision via strata-vision — mtmd already has glm5v upstream (llama.cpp#27773); ggml pin predates it

  VecSzn/PolyStrata#14 · Filter by author WildeTechSolutions opened on Oct 12, 2026
  1

• RX 7600 8 GB (gfx1102): 24K benchmark on Qwen3.6 Q2 + Ornith Mini, and the --no-gpu-experts token check

  VecSzn/PolyStrata#13 · Filter by author iMilnb opened on Oct 12, 2026
  3

• UI improvements - Add swipe/regenerate controls to the built-in chat interface

  VecSzn/PolyStrata#12 · Filter by author Ziomd opened on Oct 12, 2026
  3

• **GLM 5.3 Flash 3.0BIT on Windows — RTX 5090 D v2 24 GB + 128 GB RAM: first Windows report; context extends to 1048576 via config edit, decode stays ~24-30 tok/s, exact recall verified with filled contexts up to 500K tokens**

  VecSzn/PolyStrata#10 · Filter by author biurquezjoselina-design opened on Oct 12, 2026
  1

• [FR] Add Q5 and Q6 to Gemma 4 and Qwen 3.6

  VecSzn/PolyStrata#6 · Filter by author MaximS opened on Oct 11, 2026
  3

• DeepSeek and MiMo Flash as alternative to GLM Flash

  VecSzn/PolyStrata#4 · Filter by author TomLucidor opened on Oct 10, 2026
  5

• Would including LFM2.5 or Ling be good?

  VecSzn/PolyStrata#1 · Filter by author TomLucidor opened on Oct 10, 2026
  2