AI-powered photo cleanup that runs entirely on your Mac. Uses LiquidAI LFM to understand your photos, find the junk, and help you reclaim storage without sending a single image to the cloud.

LiquidAI fuses cutting-edge AI models with blockchain data to deliver real-time insights, predictive analytics, and ecosystem intelligence. Built for Solana and the future of decentralized innovation.

A private, local AI-powered image analysis application using LiquidAI's LFM2-VL models. All processing happens locally - no data sent to external servers.

ComfyUI nodes for Batch Image Captioning + various Vision-Language (VL) models, including InternVL3.5, Xiaomi MiMo-VL, LiquidAI LFM2-VL, Kwai Keye-VL, AIDC-AI Ovis2.5 and Ovis-U1. Models: Ovis2.5-2B, Ovis2.5-9B, Keye-VL-8B-Preview, MiMo-VL-7B-RL-GGUF, LFM2-VL-450M, LFM2-VL-1.6B, Ovis-U1-3B, Ovis2.5-2B, Ovis2.5-9B, InternVL3_5-1B/2B/4B/8B/14B/38B

Fine-tuning of LiquidAI LFM2-VL vision-language models on food image classification using LoRA

No description available

Fine-tuning of LiquidAI/LFM2.5-1.2B-Base to behave as a grumpy Italian chef, using Supervised Fine-Tuning (SFT) followed by Direct Preference Optimization (DPO)

LFM2-Audio-1.5B is Liquid AI's first end-to-end audio foundation model. Designed with low latency and real time conversation in mind

Mirror of LiquidAI/LFM2-1.2B from HuggingFace (model files excluded)

FastAPI service for Speech-to-Text, Text-to-Speech, and Conversation using LiquidAI LFM2-Audio-1.5B

Image analyzer using LFM2-VL models from LiquidAI.

Introducing liquid glass design into SaaS Webpage.

LiquidAirPlant is a MATLAB code allowing its users to model various liquid air power plant configurations driven by natural gas combustors, parabolic trough solar collectors, and/or ambient air.

This app is powered by Liquid AI's LFM-2.5-Audio-1.5B and LFM-2.5-VL-1.6B models, supporting multi-TTS functions, all offline via Gradio. Tested smoothly on RTX 5080 and functional on RTX 3060 (with occasional speech-to-speech pauses); requires 16GB to 32GB VRAM for real-time speech-to-speech to work.

Parameter-efficient fine-tuning of LiquidAI LFM2-1.2B using LoRA

Fine-tuning LiquidAI/LFM2-VL-1.6B in Colab (LoRA/4-bit) + dataset template + probe test.

Local GPU inference example using LiquidAI/LFM2-ColBERT-350M model for document retrieval and ranking

**MedAgent: A medical AI assistant powered by LiquidAI LFM2.5-1.2B-Instruct for clinical conversations with context memory and structured responses.**

Anywhere-LFM est une application de bureau légère, pensée pour simplifier l'utilisation en local des modèles LiquidAI/LFM2 (350M à 1.2B). Elle permet à n'importe qui de discuter avec une IA générative performante, sans nécessiter de ligne de commande ou de configuration complexe.

A list of Edge optimized Ollama models

Liquid-Granite is a lightweight, privacy-first implementation of the Researcher–Critic pattern that runs entirely offline via Ollama. It pairs IBM Granite-4.0 as the generator with LiquidAI LFM2.5-1.2B-thinking as the critic to create self-correcting, iteratively refined AI responses on edge hardware.

⚡🗼Liquid AI x W&B x Lambda Hackathon —🥈 2nd Place 🧠 Building creative, human-centered AI experiences. 💻 Focused on multimodal LLMs, voice interaction & fast prototyping (vibe coding). 🌸 Exploring the intersection of Japanese culture and AI innovation.

LiquidAI fuses cutting-edge AI models with blockchain data to deliver real-time insights, predictive analytics, and ecosystem intelligence. Built for Solana and the future of decentralized innovation.

No description available

No description available

A ChatBot trained with LiquidAi

LiquidAILocal

No description available

LiquidAI LFM in MLX

No description available