AI BriefingLiquid AI Releases LFM2.5-8B
·2026.05.29 13:17
Key point
Liquid AI has released LFM2.5-8B-A1B, an edge model with a 128K context and 38T tokens of training data.
Details
Liquid AI has released LFM2.5-8B-A1B, an edge model designed to power real-life applications. Compared to the existing LFM2-8B-A1B model, the following major upgrades have been applied.
- Extended 128K context window
- Significantly increased pretraining data volume (38T tokens, up from 12T)
- Large-scale Reinforcement Learning (RL) applied
- 2x expanded vocabulary for improved tokenization of non-Latin languages
This model excels at tool calling chaining and performing complex tasks, and has been optimized to run smoothly even on budget laptops.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.