Liquid AI releases LFM2.5-24B-A2B for local agents
· KGAM.Blog
On October 13, 2026, Liquid AI introduced the open-weight LFM2.5-24B-A2B. It has 24 billion parameters, with 2 billion active per token, and is designed for reasoning and tool calls.
A handy generator of ready-to-use prompts
Context expands to 131,000 tokens. Launch support includes llama.cpp, MLX, vLLM, SGLang and ONNX; both the main and base models are published.

Two billion active parameters does not mean the entire model file is that small. Check memory and speed on your own device and give the agent only the permissions it needs.