antirez's DwarfStar runs Kimi K3.1 PRO locally from safetensors

antirez's DwarfStar project can now run Kimi K3.1 PRO locally by streaming safetensors weights from SSD, generating about 1 token/second for interactive use, requiring at least 128GB of RAM.

2026-11-08 ~ 2026-11-08 · 4 related posts

• antirez: DwarfStar now runs Kimi K3.1 PRO inference from safetensors via SSD streaming — antirez · 2026-11-08
• DwarfStar streams Kimi K3.1 PRO from safetensors via SSD for local inference — Goodnesmbakara · 2026-11-08
• antirez: Kimi K3.1 runs locally at 1 token/sec, good enough for interactive QA — antirez · 2026-11-08
• antirez: 1 token/sec local Kimi K3.1 inference needs 128GB memory, 512GB much better — antirez · 2026-11-08