llama.cpp adds Sparse MoE support, Qwen3.6 GGUF, and WebWorld models for local AI

The llama.cpp project has been updated to support Xiaomi's MiMo-V2.5 Sparse MoE model, allowing local inference of large, parameter-efficient models. Additionally, a new uncensored Qwen3.6 27B model is now available in GGUF format for local use, featuring improved performance and fewer refusals. The WebWorld series, based on Qwen3, has also been released with multiple parameter sizes to facilitate the development of local web agents capable of interacting with online environments. AI

Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →

IMPACT Enhances local AI capabilities by enabling more efficient inference of advanced MoE models and providing specialized models for web agent development.

RANK_REASON This cluster details updates to open-source inference engines and the release of new open-weight models, fitting the research category. [lever_c_demoted from research: ic=1 ai=1.0]

Read on dev.to — LLM tag →

• llama.cpp
• Xiaomi
• MiMo-V2.5
• Qwen3.6
• GGUF
• WebWorld
• Qwen3
• Ollama

• model release
• product