Liquid AI Launches New Fast Encoders
Liquid AI has launched two new models called LFM2.5-Encoder-230M and 350M. These tools are designed to handle long texts of up to 8,000 tokens while staying very efficient.
The larger 350M version shows great results on language benchmarks. It ranks near the top compared to much bigger models on a variety of testing tasks.
These models are built to run fast on your own computer. The 230M version can process a full 8,000-token document on a CPU in under half a minute.
Comments (0)
No comments yet. Be the first!
More AI news
NewsAI Agents Help Scientists Build Software Faster
A new report shows that AI coding assistants are speeding up scientific software development.
NewsAI Now Fuels One in Four Cyberattacks
Hackers are increasingly using AI to carry out costly data breaches.
NewsThe Danger of Vibe Lawyering
Relying on AI to handle legal work without checking the facts is leading to serious errors in court.