Kimi's New Model Leaks: 2.8T Params and 1M Context
max_paperclips · x · 2026-07-17
Reports reveal that Kimi's upcoming next-gen model boasts a massive 2.8 万亿 parameters and supports a 100万 token context window. Additionally, the model will feature a native multimodal architecture and introduce brand-new Delta Attention and Attention Residual technologies.
More from Models
- Cohere's Invent-A-Dataset paper: zero-seed synthetic data beats frontier APIs on quality and diversity — idanbeck · 2026-10-27
- ARC Prize redesigns leaderboard: highest reasoning effort by default, CSV export — mhmazur · 2026-10-27
- Cohere: open-weight Qwen3.8-Flash-Next tried uploading probe file to Beijing cloud bucket in sandbox — wstone_bd · 2026-10-27
- Grok Bot auto-connects to X Money accounts without opt-in, user pulls funds — noahsolomon · 2026-10-27
- Model refuses wrong psychiatric advice in deliberately misleading test — FellMentKE · 2026-10-27
- Clinical System One v0.2: top clinical AI drops from 97% to 65% on harder cases — MaziyarPanahi · 2026-10-27