Liquid AI Launches LFM3 Model Family; 2B Vision Model Runs Live Camera on iPhone Offline
JosephJacks_ · x · 2026-11-03
Liquid AI introduced LFM3, a new generation of Liquid Foundation Models with a tri-hybrid architecture and multimodal encoders, spanning 6 instances: LFM3-1.7B (text, 256k context), LFM3-VL-2B (vision-language with video input), LFM3-Audio-2B (tool calls, voice chat, 18 languages), LFM3-Omni-2B, plus d1-2B and d1-Omni-2B decision models with calibrated probabilities.
A developer demoed LFM3-VL-2B in the iOS app Local LLM, running fully offline (airplane mode) live camera understanding and voice chat on an iPhone, showing what a 2B model can do on-device.
More from Models
- GPT-6 edits 0.016% of Qwen3.5-0.8B weights, boosting math scores up to 6x with no training — SonglinYang4 · 2026-11-03
- Claude's alignment answers aren't literal defaults but a posture of supplying the right answer — 1a3orn · 2026-11-03
- Claude models echo Anthropic's stance on alignment questions even with reworded prompts — 1a3orn · 2026-11-03
- OpenAI's new Daybreak mode still blocks cybersecurity requests like code injection — yassineyousfi_ · 2026-11-03
- User flags a 'concerning' pull request generated by Opus 5.5 — Sauers_ · 2026-11-03
- ChatGPT's browser refuses to look up Hitler videos — 'I thought this was a web browser' — JasonBotterill · 2026-11-03