Library
85,762 resources from 11 public sources. 53 checked by hand; the rest are unreviewed.
Results
-
MiniCPM-OCR-Jev
Modelsimage text to text model by openbmb · transformers
openbmb/MiniCPM-OCR-Jev -
WeVisOne-VL-4B
Modelsimage text to text model by tencent · transformers
tencent/WeVisOne-VL-4B -
LightOnOCR-3-0.8B
Modelsimage text to text model by lightonai · transformers
lightonai/LightOnOCR-3-0.8B -
NV-Reason-CT
Modelsimage text to text model by nvidia · transformers
nvidia/NV-Reason-CT -
CaPtain
Agents & frameworksCaPtain: A Pipeline for Post-Training Code-as-Policy Robot Agents — CaPtain-Harness, SFT and RL training code, benchmark adapters
git clone https://github.com/waltstephen/CaPtain.git -
Mage-VL-Cowork
Modelsimage text to text model by microsoft · transformers
microsoft/Mage-VL-Cowork -
WeVisOne-VL-2B
Modelsimage text to text model by tencent · transformers
tencent/WeVisOne-VL-2B -
video-search-and-summarization
AI repositoriesNVIDIA 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.
git clone https://github.com/NVIDIA-AI-Blueprints/video-search-and-summarization.git -
humanities_data_benchmark
Datasets & benchmarksLLM Benchmark Suite for Humanities Data
git clone https://github.com/RISE-UNIBAS/humanities_data_benchmark.git -
LightOnOCR-3-1B
Modelsimage text to text model by lightonai · transformers
lightonai/LightOnOCR-3-1B -
model by lightx2v · minimax-h3
lightx2v/MiniMax-H3-Prompt-Rewriter-Qwen3.8-27B-Ref2VA -
vllm-mlx
AI repositoriesHigh-performance OpenAI and Anthropic compatible LLM inference server for Apple Silicon. Native MLX, continuous batching, multimodal models, MCP tool calling, and Claude Code support.
git clone https://github.com/waybarrios/vllm-mlx.git -
kanana-2-v-decide-3b-2611
Modelszero shot classification model by kakaocorp · transformers
kakaocorp/kanana-2-v-decide-3b-2611 -
NEWTON
Agents & frameworksNEWTON: Agentic Planning for Physically Grounded Video Generation
git clone https://github.com/CUTEPKQ/NEWTON.git -
Xiaomi-OCR-0
Modelsimage text to text model by SeerRay-Lab · transformers
SeerRay-Lab/Xiaomi-OCR-0 -
HallusionBench
Datasets & benchmarks[CVPR'24] HallusionBench: You See What You Think? Or You Think What You See? An Image-Context Reasoning Benchmark Challenging for GPT-4V(ision), LLaVA-1.5, and Other Multi-modality Models
git clone https://github.com/tianyi-lab/HallusionBench.git -
Cradle
AI repositoriesThe Cradle framework is a first attempt at General Computer Control (GCC). Cradle supports agents to ace any computer task by enabling strong reasoning abilities, self-improvment, and skill curation, in a standardized general environment with minimal requirements.
git clone https://github.com/BAAI-Agents/Cradle.git -
CT-RATE
Datasets & benchmarksHugging Face dataset by ibrahimhamamci
ibrahimhamamci/CT-RATE -
ChartVerse-SFT-1.8M
Datasets & benchmarksHugging Face dataset by opendatalab
opendatalab/ChartVerse-SFT-1.8M -
wechat-mac-rpa
Agents & frameworks基于多模态视觉感知与 LLM Agent 的 macOS 微信自动化框架 | Visual RPA for WeChat
git clone https://github.com/wq19901103wq/wechat-mac-rpa.git -
Alpie_seed
Modelsimage text to text model by 169Pi · transformers
169Pi/Alpie_seed -
MiniCPM-V-4.7-1B-heretic
Modelsimage text to text model by saidutta69 · transformers
saidutta69/MiniCPM-V-4.7-1B-heretic -
Awesome-CUA
Agents & frameworksOpen, source-backed catalog of computer-use agents: reviewed papers, GUI models, datasets, benchmarks, tools, and reusable JSONL/SQLite data.
git clone https://github.com/reacher-z/Awesome-CUA.git -
dsh-vision-toolkit
Skills[dsh]为纯文本模型设计更强大的视觉工具箱:一行安装使用、粘贴图片直接识别、多张图片问答、截图到前端UI 还原等|DeepSeek Harness-native integration for agent-vision-toolkit: image Q&A, long-screenshot OCR, UI restoration, grounding, pixel diff, Artifacts, and Web UI.
git clone https://github.com/Anionex/dsh-vision-toolkit.git
Indexed entries aren't reviewed or tested by InnerVault. Details come straight from each source. Anything that can reach your accounts, files or browser can do damage, so read the source before you install it.