Havuz

Stop paying the multimodal tax. Your AI agent keeps every image it has seen but only sends the ones it needs: lighter chats, no endless reco

Search your computer by what's in your files: photos, PDFs, documents, recordings, videos and code. On-device on Mac, Windows and Linux, bui

A native .NET LLM inference engine and agent runtime for GGUF models. TensorSharp provides a console application, a web-based chatbot interf

Official Code for OpenVLHarness: Real-World Visual Agentic Tool‑Use with Multimodal Memory

🚀🚀 A multimodal System One decision model that gives calibrated answers to typed questions about screens, photos, video and text in one forw

This project is for learning: it shows how an agent loop works underneath, in a way more intuitive and visual than reading about it. Behind

PixelRAG, web sayfaları ve belgeleri screenshot tile’larına dönüştürüp görsel içerik üzerinden arayan bir RAG sistemi sunuyor.

Neden yükseliyor · 121 günde 10.101 yıldıza ulaşması ve 8,28 milyon Wikipedia sayfasını içeren canlı indeks/API sunması dikkat çekiyor. README, Berkeley SkyLab, BAIR ve Berkeley NLP bağlantılı araştırma çalışmasını resmi kod tabanı olarak tanımlıyor. Son 28 günde 5 commit ve 4 farklı yazar görünmesi katkı çeşitliliğine işaret etse de aktiviteyi çok yüksek göstermiyor.

AI-enabled Intelligent Document Processing for Python: 12+ LLM backends, Pydantic schemas, self-hosted OCR (Nanonets-OCR2-3B), auto-schema d

Modality-First GenAI Platform

A catalog of every MLP neuron and attention head of Qwen3-8B and Gemma 4 12B, read from the weights through two Jacobian lenses (R and S), f

Halo is an open-source framework built by White Circle for training large language and multimodal models

DeepSeek Harness (DSH) plugin: dispatch work to DSH agents from Claude Code / Codex — native subagent progress, in-host worker sessions with

An open-source, agent-native platform for multimodal AI generation, built on DeepSeek Harness. Every interaction lives in agent conversation

A Codex plugin and local MCP runtime for context-bounded image generation and inspection.

claude-real-video, video dosyası veya URL’den sahne değişimlerine göre seçilmiş kareler ve transcript çıkarıp bunları LLM’lerin inceleyebileceği yerel çıktılara dönüştürür.

Neden yükseliyor · Video analizini mevcut LLM/agent iş akışlarına ekleyen kurulumu kolay bir araç sunuyor; README’de Hacker News ön sayfa rozeti var. 89 günde 2.185 yıldız ve son 28 günde 16 commit ile 4 farklı yazar, ilginin yanında yakın dönem katkı etkinliğine de işaret ediyor; yıldızlar tek başına kullanım veya kalite kanıtı değil.

Human-feedback and data-governance infrastructure for RSI research: annotation, review, provenance and frozen dataset candidates.

On-device, real-time multimodal AI with features similar to GPT-Live

Give Claude the ability to watch and understand videos — Claude Code plugin with frame extraction and multimodal audio analysis

Give AI agents eyes, ears, and verifiable results. Watch Skill turns video, audio and screen activity into searchable, timestamped evidence

LLMInternSkill: LLM internship resume and job-search Codex Skill for resume polish, JD tailoring, evidence guard, interview grilling, and Pr

Create high-fidelity Codex companions from 2+ photos with local open-weight neural editing and no OpenAI API key.

OpenCode skill suite for MiniMax H3 directing, routing, multishot planning, prompt generation, and review.

A Codex plugin and local MCP runtime for context-bounded image generation and inspection.

Cheap Money-Saving Self-hosted AI agent workspace with tool calling, MCP, multi-model routing, sandboxed execution, multi-agent workflows, a

为 Claude Code 赋能多模态视觉能力，适配 纯文本 LLM 底座，用于截图 / UI / 图表分析；搭配 browser-harness 可做前端布局自动化检查。

Open-source video processing core and CLI. Search and annotate local videos with your own model endpoints.

Production-oriented Spring AI platform prototype for RAG, tool calling, async ingestion, JWT/RBAC, and observability.

The fastest way to put Volcengine Ark in your terminal and your AI agent — go from prompt to generated media, multimodal answer, or deploy

NEWTON: Agentic Planning for Physically Grounded Video Generation

Learning Hierarchical and Geometry-Aware Graph Representations for Text-to-CAD

World's fastest and most compact embedded vector database: exact by default, multimodal, local-first, and GPU-accelerated

The vision plugin for OpenCode that truly understands images. Inspect, read, and reason about any screenshot or picture with deeper understa

一条命令，把微信变成任何 AI Agent 的入口 | Connect WeChat to any AI Agent with one command

Transparent proxy for Claude Code that auto-routes image-bearing requests to a multimodal model — so a non-multimodal primary model never cr

Agentic, long-horizon visual generation: a fuzzy story → a cross-model-audited image-based movie. Brings ARIS's research-wiki + multi-agent

Self-hosted, private family AI health steward. Multimodal AI structures health data into a person-centric profile with dashboard & AI consul

专利侵权分析系统 —— 输入专利公开号，产出竞品侵权分析报告；同时打包成 skill，可被任意 agent（dsh, codex, claudecode 等） 调用。

Open-source, customizable frontend for Venice AI. Chat, image gen, audio, video, embeddings + visual workflows — all in one UI. Your API key

给 DeepSeek 补上「眼睛和耳朵」的多模态视觉插件：看图 / OCR / 物体检测 / 视频理解 / 语音转写 / 截图直读，一键安装（DSH 插件）。

Blazing fast Go port of gemini-web2api. Convert Google Gemini web into OpenAI-compatible API. Zero cost, single static binary.

让任何 AI 真正"看"视频：纯本地的视频阅读 skill——faster-whisper GPU 转写 + ffmpeg 场景感知抽帧，双通道带时间戳（t=MM:SS），数据不出机。1 小时视频 3~5 分钟读完。Give any AI agent eyes for video

An agent that edits your Word and Excel files — then looks at them, through real Microsoft Office, to check its own work

Source-linked collection of published AI model benchmark scores.

基于 Spring Boot、微信 iLink SDK 与阿里云百炼的多模态微信机器人，支持文本、图片、语音、Function Calling、Skill 编排和本地 RAG。

Pixel-native visual RAG ported to Rust on the ruvector ANN substrate (HNSW + IVF-Flat) — screenshot/document retrieval over visual embedding

MODA: open fashion retrieval benchmark and models by Hopit AI. MODA (203M, open source), MODA Pro Lite (213M, open weights), MODA Pro (hoste

A token-efficient, provider-agnostic vision layer for reasoning models via MCP.

Official code for InfoVis: one 9B model for slide image generation and editing, with the VisSlideBench benchmark and 15M+ training records.

[MM' 2026] Light-MER for efficient multimodal emotion recognition with sub-1B MLLMs.