LiquidAI launches two multilingual encoder models with CPU speed and long-context gains
maximelabonne · x · 2026-07-28
LiquidAI released two new encoder models, LFM2.5-Encoder-230M and LFM2.5-Encoder-350M, and says they are fast to train, work well on long context, and are strongly multilingual.
The company highlights a demo set on Hugging Face and cites the 230M model as about 3.7x faster than ModernBERT-base on CPU at 8,192 tokens. The 350M model ranks 4th of 14 on GLUE, SuperGLUE, and multilingual classification, behind only three larger models, one of them nearly 10x its size.
Related event: Liquid AI releases LFM2.5 Encoder 230M and 350M(7 posts)→
More from Models
- NVIDIA open-weight model ties pplx-decider at #1 on 669-case clinical benchmark — MaziyarPanahi · 2026-10-16
- Claude Code adds idle compaction to save 1h prompt cache on idle sessions — EricBuess · 2026-10-16
- Thrive Holdings launches real-work AI evals from 75+ businesses, benchmarking 13 models — samaysham · 2026-10-16
- Microsoft's MAI-Transcribe now handles 61% of all transcription on OpenRouter — mustafasuleyman · 2026-10-16
- Reflection AI Releases Open-Weight 501B MoE Beam for Agentic Coding on AWS — AWS ML Blog · 2026-10-16
- Dev RL-tunes Qwen 27B against slop with Meta's XAR method, lands top-4 on EQbench — Mysterious_Hearing14 · 2026-10-16