Is the 342GB Pruned and Quantized Kimi K3 Usable?
Hannibalj2ca · reddit · 2026-07-30
A developer has released a REAP55 pruned and IQ1M quantized version of the Kimi K3 model on Hugging Face, weighing in at a massive 342GB. The poster is asking the community if anyone has tested this build and how it performs in terms of actual usability and speed.
More from Models
- Trying Grok bot surfaces an unpaid Cursor balance, blocking account merging — haydendevs · 2026-10-16
- TensorFold 1.0.12 runs 120B vision model on 1-2 DGX Sparks, adds Mac image input — HankYeomans · 2026-10-16
- Leaked screenshot shows unannounced Gemini 4 Argon gated to Ultra and paid Pro subscribers — il_-bro · 2026-10-16
- Google: Gemini 4 Argon lands in Antigravity, code submissions up 35%, rollbacks down ~33% — tekbog · 2026-10-16
- Google's "ungrounded beliefs" classifier injections derail Opus 4.1's reasoning chain — liminal_bardo · 2026-10-16
- A week of ChatGPT Pro quota burned in 36 hours as GPT-6 fails to factor a 512-bit RSA key — matthew_d_green · 2026-10-16