Moonshot AI detailed how it trained its Kimi K3 model, moving from architecture to a unified system through pretraining, SFT, reinforcement learning, and MOPD. The company trained the MoonViT-V2 visio…
Moonshot AI released Kimi K3, a 2.8-trillion-parameter open-weight model, the largest open-source model as of mid-2026 and the first to reach the 3-trillion-parameter class. The model uses Kimi Delta …
Moonshot AI's Kimi K3 model scales to 2.8T total parameters with 104B activated per token, using hybrid attention, Attention Residuals, and Stable LatentMoE to support 1M-token agentic trajectories. T…
A developer released a PyTorch implementation of the Kimi K3 architecture from the arXiv paper 'Kimi K3: Open Frontier Intelligence' (arXiv:2607.24653v1), reproducing the paper's Table 1 parameter cou…
Moonshot AI released the full technical architecture of its Kimi K3 model, a 2.8 trillion-parameter multimodal mixture-of-experts system with a one-million-token context window, on July 27. The 1.56 T…
Moonshotai released Kimi-K3, a 2.8 trillion parameter open-weight multimodal agentic model on Huggingface, claiming it is the world's first open 3T-class model. It features a 1-million-token context w…
Moonshot AI released Kimi K3, an open-weight 2.8-trillion-parameter native multimodal agentic model with a 1-million-token context window, claiming it is the world's first open 3T-class model. Built o…