Liquid AI, an MIT spin-off, has released LFM2.5-24B-A2B, an open-weight model with 24 billion parameters that activates only 2 billion per token. The model reaches up to 184 tokens per second on Apple’s M5 Max chip and runs up to 1.7 times faster than comparable mixture-of-experts models on edge devices. It leads or ties with larger models on 5 out of 8 agentic and tool-use benchmarks, including a standout score of 27.69 on AutomationBench versus 8.72 for Qwen3.6-35B-A3B. The model features a 131,000-token context window, a 128,000-entry vocabulary, and supports 16 languages. It was released on October 13, 2026 and is available on Hugging Face.
Source: Read the original article
Disclaimer: this content is for information purposes only and is not financial advice. Cryptocurrencies are highly volatile: you may lose all of your capital. Always do your own research. Legal notice
