Liquid AI ships LFM3: six 2B-class edge models spanning text, vision, audio, omni
JosephJacks_ · x · 2026-10-28
Liquid AI released LFM3, a six-instance model family (text, vision-language, audio, omni, plus two decision models) built on a new tri-hybrid architecture. Highlights: LFM3-1.7B with 256k context, LFM3-VL-2B claimed to match GPT 5.6 Luna / Gemini Flash Ultra, d1 decision models with calibrated probabilities in one pass. Architecture was auto-searched against 14 processor types, targeting the fragmented edge-device world with cheap fine-tuning and fast deployment.
Related event: LiquidAI Releases LFM3 Family of On-Device Models(15 posts)→
More from Models
- Screenshot Appears to Show GPT-6 and GPT-5.6 Sol in ChatGPT Model Picker — ShelterCultural5725 · 2026-10-28
- ChatGPT Work sync bug makes messages vanish from view while the model still recalls them — LynxAirSound · 2026-10-28
- Palisade Research finds Alibaba's Qwen 3.8 models routed some requests to company servers — The Decoder · 2026-10-28
- DeepSeek on $75K of hardware: 2,120 tok/s decode at 16 concurrent sessions — victormustar · 2026-10-28
- Frontier models hit 73.3% on SpatialBench 10 months after release, near 75% ceiling — kenbwork · 2026-10-28
- 50-task Salesforce agent eval: GPT-6.1 Sol scores 72%, well behind Claude Opus at 87% — TruthDeep8659 · 2026-10-28