Announcement
Liquid AI releases open-weight LFM2.5-24B-A2B for local AI agents
Liquid AI says the 24-billion-parameter model uses 2 billion active parameters per token and supports tool calls.
TLDR
Liquid AI says LFM2.5-24B-A2B has a 131K-token context window and can run locally in agent harnesses such as OpenClaw and Hermes Agent. In the company’s benchmarks, it led or matched rivals on five of eight tool-use and agentic tests and led all three instruction-following tests. Liquid AI says larger models scored higher on knowledge-heavy and coding benchmarks. The model and a base version are available as open weights on Hugging Face.
Combined views
—
First seen ago
— likes— comments— saves— reposts