RobotecAI / rai

RAI is a vendor agnostic agentic framework for Physical AI robotics, utilizing ROS 2 tools to perform complex actions, defined scenarios, free interface execution, log summaries, voice interaction and more.

rockmatrix-ai / EdgeRock

EdgeRock: Strong Reasoning at Constant Memory. Multimodal models with O(N) generation and O(1) persistent history. Inference and evaluation code.

modal-labs / modal-client

SDK libraries for Modal

opengeos / GeoAgent

A multimodal AI agent for geospatial data analysis and interactive visualization

oxbshw / watch-skill

Give AI agents eyes, ears, and verifiable results. Watch Skill turns video, audio and screen activity into searchable, timestamped evidence and proves work with deterministic contracts, not model opinion. DeepWatch is the agent workspace built on DeepSeek Harness. Python + npm, MCP, CLI, REST, Web.

PacesaLab / BindCraft2

Easy to use suite for designing a wide range of protein binder modalities.

PipeNetwork / kimi-k3-mlx

MLX port of moonshotai/Kimi-K3 (2.78T multimodal MoE): streaming converter, REAP expert pruning, and per-language expert-overlap analysis

duckyshell / ComfyUI-MiniMaxH3-Prompt-Writer

Multimodal MiniMax H3 prompt writer for ComfyUI with local and API providers.

chflame163 / ComfyUI_Qwen_H3_Prompt

Use local Qwen3.8-27B and official MiniMax-H3 Skills in ComfyUI to generate H3 prompts. Runs a standalone local llama-server , inference operates fully offline with no internet access.

czvvd / PartLLM

[SIGGRAPH Asia 2026 (TOG)] PartLLM: A Unified Multimodal Foundation for 3D Part Segmentation

humza-khalid / 12vhpwr-guard

Standalone per-pin 12VHPWR overcurrent watchdog for ASUS ROG Astral GPUs. Shuts down your PC before the connector melts.

SimpleJev / JevAny

Open infrastructure for training, evaluating, and deploying System 1 decision models across language and multimodal backbones.

t3knobox / klik-anim-skill-creation

Claude skills for industry-standard Remotion motion graphics: klik-anim-bible (animation craft laws, standalone) + klik-remotion-method (end-to-end video production pipeline). Install one or both — see README.

dashs20 / CRAP

CRAP: Cycloidal Reducer Algorithm & Parametrizer. Parametric cycloidal gearbox generator (build123d): give it bearings, motor, ratio and eccentricity; get collision-checked, toleranced STEP/STL parts and a live 3D web GUI.

Roclp / TI2V-Jailbreak

Code for **"The Price of Consistency: Exploiting Visual Anchors for Multimodal Jailbreaking in Video Generation"** (ACM MM 2026).

DeepExperience / VAD_Multimodal_OPD

—

johngarciahq19v3 / Helldivers-2-Ultimate-Assistant

Ultimate assistant for Helldivers 2. Features: God Mode, Infinite Ammo, Infinite Grenades, Infinite Stims, No Reload, Unlimited Stratagems, Super Credits & Medals Farm, Infinite Mission Timer, and more. 20+ functions.

PsiACE / dohnuts

Dohnuts builds small multimodal models for direct decisions. -> System One model

davidalmeida90 / quant-models

Open-source quantitative finance models in Python: option pricing and Greeks, volatility surfaces, deep hedging, portfolio construction and Monte Carlo. Every model runs on free data, with the full write-up on davidariasfinance.com

microsoft / unilm

Large-scale Self-supervised Pre-training Across Tasks, Languages, and Modalities

TEN-framework / ten-framework

Open-source framework for conversational voice AI agents

yzhao062 / pyod

A Python library for anomaly detection across tabular, time series, graph, text, image, and audio data. 60+ detectors, benchmark-backed ADEngine orchestration, and an agentic workflow for AI agents.

xorbitsai / inference

Swap GPT for any LLM by changing a single line of code. Xinference lets you run open-source, speech, and multimodal models on cloud, on-prem, or your laptop — all through one unified, production-ready inference API.

modelscope / modelscope

ModelScope: bring the notion of Model-as-a-Service to life.

voila-dashboards / voila

Voilà turns Jupyter notebooks into standalone web applications

OpenBMB / UltraRAG

A Low-Code MCP Framework for Building Complex and Innovative RAG Pipelines

facebookresearch / mmf

A modular framework for vision & language multimodal research from Facebook AI Research (FAIR)

InternLM / xtuner

A Next-Generation Training Engine Built for Ultra-Large MoE Models

ParisNeo / lollms-webui

Lord of Large Language and Multi modal Systems Web User Interface

phodal / awesome-iot

Awesome IoT. A collaborative list of great resources about IoT Framework, Library, OS, Platform

kyegomez / tree-of-thoughts

Plug in and Play Implementation of Tree of Thoughts: Deliberate Problem Solving with Large Language Models that Elevates Model Reasoning by atleast 70%

EvolvingLMMs-Lab / lmms-eval

One-for-All Multimodal Evaluation Toolkit Across Text, Image, Video, and Audio Tasks

open-compass / VLMEvalKit

Open-source evaluation toolkit of large multi-modality models (LMMs), support 220+ LMMs, 80+ benchmarks

aurelio-labs / semantic-router

Superfast AI decision making and intelligent processing of multi-modal data.

daviddrysdale / python-phonenumbers

Python port of Google's libphonenumber

morphik-org / morphik-core

Open-source multimodal retrieval engine (Morphik Core)

embeddings-benchmark / mteb

MTEB: State-of-the-art evaluation of embeddings across languages and modalities

nicedreamzapp / claude-code-local

Run Claude Code 100% on-device with local AI on Apple Silicon. MLX-native Anthropic-API server. 6 fighters incl. Muse-Glimmer 30B (now multimodal — reads images, abliterated), Gemma 4 31B, Qwen 3.5 122B (65 tok/s), DeepSeek V4 Flash (1M ctx). Private, offline, airgap-ready. Built for NDA / legal / healthcare workflows.

beeware / briefcase

Tools to support converting a Python project into a standalone native application.

dalibo / pg_activity

pg_activity is a top like application for PostgreSQL server activity monitoring.

datachain-ai / datachain

The Context Layer for unstructured data: typed, versioned datasets over S3, GCS, Azure

autodistill / autodistill

Images to inference with no labeling (use foundation models to train supervised models).

roboflow / maestro

streamline the fine-tuning process for multimodal models: PaliGemma 2, Florence-2, and Qwen2.5-VL

om-ai-lab / OmAgent

[EMNLP-2024] Build multimodal language agents for fast prototype and production

rasterio / rasterio

Rasterio reads and writes geospatial raster datasets

zai-org / GLM-V

GLM-4.6V/4.5V/4.1V-Thinking: Towards Versatile Multimodal Reasoning with Scalable Reinforcement Learning

ByteDance-Seed / VeOmni

VeOmni: Scaling Any Modality Model Training with Model-Centric Distributed Recipe Zoo

google-gemini / genai-processors

GenAI Processors is a lightweight Python library that enables efficient, parallel content processing.

fabiocaccamo / django-admin-interface

:superhero: :zap: django's default admin interface with superpowers - customizable themes, popup windows replaced by modals and many other features.

shcherbak-ai / contextgem

Extract structured data and insights from documents using LLMs and decision models

Yuliang-Liu / Monkey

Monkey (LMM): Image Resolution and Text Label Are Important Things for Large Multi-modal Models (CVPR 2024 Highlight)

kyegomez / BitNet

Implementation of "BitNet: Scaling 1-bit Transformers for Large Language Models" in pytorch

NVIDIA-AI-Blueprints / video-search-and-summarization

NVIDIA AI Blueprint for video search and summarization (VSS) is a GPU-accelerated reference architecture for building video analytics agents with real-time verified alerts, visual Q&A, and automated reporting. The VSS Blueprint uses vision language models (VLMs) such as NVIDIA Cosmos, LLMs such as NVIDIA Nemotron, RAG, and NVIDIA NIMs.

Kav-K / GPTDiscord

A robust, all-in-one GPT interface for Discord. ChatGPT-style conversations, image generation, AI-moderation, custom indexes/knowledgebase, youtube summarizer, and more!

Tencent / WeMM-Embedding

WeMM-Embedding is a family of universal multimodal embedding models by the WeChat Vision Team at Tencent, supporting multimodal understanding and retrieval.

facebookresearch / multimodal

TorchMultimodal is a PyTorch library for training state-of-the-art multimodal multi-task models at scale.

HumanSignal / Adala

Adala: Autonomous DAta (Labeling) Agent framework

pixeltable / pixeltable

The backend agents build with - Multimodal database, orchestration, and serving in one file

autonomousvision / transfuser

[PAMI'23] TransFuser: Imitation with Transformer-Based Sensor Fusion for Autonomous Driving; [CVPR'21] Multi-Modal Fusion Transformer for End-to-End Autonomous Driving

vturrisi / solo-learn

solo-learn: a library of self-supervised methods for visual representation learning powered by Pytorch Lightning