Moonshot puts 2.8T-parameter Kimi K3 on Hugging Face with 1M-token context
DeryaTR_ · x · 2026-07-28
Moonshot’s Kimi K3 is now on Hugging Face as an open-weight model.
- Scale: 2.8T parameters, advertised as the world’s first open 3T-class model.
- Architecture: built on Kimi Delta Attention (KDA), Attention Residuals (AttnRes), and a Stable LatentMoE setup that activates 16 of 896 experts.
- Capabilities: native vision, a 1M-token context window, and positioning around long-horizon coding, knowledge work, and reasoning.
- The model card says the release is meant for research, deployment, and further development under the Kimi K3 license.
Related event: Moonshot Open-Sources 2.8T Parameter Kimi K3(32 posts)→
More from Models
- Kimi K3 weight shard appears as `model-00001-of-000096.safetensors` — ricklamers · 2026-07-28
- Microsoft reiterates MAI-Cyber-1-Flash and MDASH as a lower-cost security stack — satyanadella · 2026-07-28
- Microsoft launches MAI-Cyber-1-Flash and MDASH, claiming top CyberGym results at half the cost — satyanadella · 2026-07-28
- Claude Opus 5 keeps Opus 4.8 pricing while matching Fable 5 within 0.5% on coding — AlexKim · 2026-07-28
- Claude Opus 5’s migration guide quietly changes years of prompting advice — AlexKim · 2026-07-28
- Post says an attention-heavy model has 104B active parameters — zephyr_z9 · 2026-07-28