Repositorystudio-dots-ai/dots3-note-prevAnnouncement
MoE architectures that activate only a small fraction of their parameters (16B of 280B) while handling 512K context windows lower the operational threshold for running high-capability, open-weight multimodal agents.
The repository introduces a large-scale open-weight model addressing multi-step reasoning, agent execution, and long-context processing. dots3-note activates only 16B parameters per token, balancing compute efficiency with broad capacity across text, audio, image, and video modalities. It leverages modern serving frameworks including vLLM and SGLang alongside standard Hugging Face tools. The project is primarily aimed at AI engineers, researchers, and enterprises building self-hosted agent stacks. Regarding maturity, it is currently positioned as a preview: the weights and base integration pull requests are live under the Apache 2.0 license, but the comprehensive technical report is still marked as forthcoming.
Summary generated by AI (Gemini 3.8 Flash, 27 September 2026). It may contain errors – the original source is authoritative.
Similar in content, found by the AI search.
Link:Moonshot AI
Moonshot AI has released Kimi K3, an open-weight native multimodal agentic model featuring 2.8 trillion total parameters and a 1-million-token context window. The architecture uses a sparse MoE setup that activates 104 billion parameters per token.
Link:Qwen
The Qwen team has released Qwen3.8-27B, a compact, dense vision-language model under the Apache-2.0 licence. The 27-billion-parameter model pairs Gated DeltaNet with conventional gated attention, handles images and videos natively, and introduces flexible reasoning controls.
Link:Xiaomi MiMo Team
Xiaomi has unveiled MiMo-V2.5-Pro, an open-source Mixture-of-Experts model featuring 1.02 trillion total parameters with 42 billion active. The model supports a 1M-token context window and is specifically designed for complex agentic workflows and long-horizon software engineering.
Link:Z.ai
Z.ai has launched GLM-5.3-Flash under the MIT licence on Hugging Face, marking the first natively multimodal model in the GLM-5 family. Featuring 320B total parameters and 18B active parameters, it targets high efficiency across coding and agentic tasks.
Repository:PipeNetwork/kimi-k3-mlx
kimi-k3-mlx is an Apple MLX port of Moonshot AI's native multimodal MoE model Kimi-K3 (2.78T total parameters). Because the model exceeds the memory capacity of any single Apple Silicon machine, the repository includes a custom streaming converter, vision components, and expert-pruning tools.
337starsPython
Link:ornith-ai
ornith-ai has released Ornith-1.5-397B, an open-source multimodal mixture-of-experts model under the MIT licence. With 397 billion parameters and an end-to-end self-improvement loop, it delivers coding and reasoning benchmark performance comparable to top proprietary models.
Parts of this content were created with the assistance of AI.
