ocr

One Mac process. Many models. Real speed. Multi-model LLM serving with prefix reuse, MTP acceleration, and OpenAI APIs — built for Apple Silicon, measured against mlx-lm and llama.cpp.

OCR system for handwritten medical prescriptions using Donut transformer and zero-shot classification

An open source framework for Retrieval-Augmented System (RAG) uses semantic search helps to retrieve the expected results and generate human readable conversational response with the help of LLM (Large Language Model).

Don't write a bug report — record it. Local MCP: screen recording → transcript, frames, OCR, wall-clock evidence for coding agents.

Desktop AI agent for Windows — an architecture that expands what any model can do. Image/video OCR, offline speech transcription, an agent-driven browser, real .xlsx/.pptx/.docx output, MCP, and a delegated team of agents. Runs on your machine, with any provider.

📋 Clipboard manager for vibe coders — native macOS menu-bar app. Text & image history, screen capture + annotation, scrolling capture, screen recording → video/GIF, OCR, on-device voice & meeting notes (or OpenAI/Gemini), share links from your own S3 bucket. Free, private, open source.

A low-token visual evidence compiler for text-only coding agents. Convert images into compact Visual Evidence Packets (VEP) for DeepSeek, Codex, Claude Code, and other text-only models.

Codex Skill：让纯文本模型（DeepSeek）借助 StepFun step-3.7-flash 获得看图能力 | Give text-only Codex models (DeepSeek) image understanding via StepFun step-3.7-flash

LegalEase AI is a document simplification tool built using Gemini API, Streamlit, and Hugging Face models. It allows users to upload legal PDFs and automatically receive simplified summaries, clause-level insights, and structured information designed for clarity and accessibility.

MCP server that gives text-only LLM coding agents vision — analyze images via any multimodal model (mimo, Claude, Gemini, OpenAI-compatible). Works with Claude Code, Codex, Kimi, opencode, PI.

DeepSeek Harness 插件：DeepSeek Pro 大脑 + 自动识图。GUI 附加图片默认经官方 deepseek-v4-flash-vision-exp 原生识图，转译成文字后交给 DeepSeek 作答（纯文本的 V4-Pro 也能看图）；支持百炼/智谱/OpenRouter 等任意 OpenAI 兼容 VLM，无 key 自动探测本地 Ollama；安装时有一问式确认

AI hardware fit calculator for LLM, embedding, reranker, OCR and VLM workloads — VRAM, throughput, licensing and multi-GPU planning

CLI, MCP server, and TypeScript library that extracts text, layout, and page images from PDFs for AI agents — when extraction goes wrong, it says so per page and says what to do next.

Token-lean web microfetch for LLM agents: any URL → clean markdown via CLI, MCP server, and Claude Code plugin. Real browser-cookie auth, passkeys, anti-bot reach, on-by-default prompt-injection defense, plus on-device multimodal ASR/OCR. A single Rust binary — not a browser.

Multi-model LLM serving for NVIDIA DGX Spark with vLLM, web UI, and tool calling

📑 npm pacakge to Craft files into Markdown with ease

Read PGS (Bluray) and VobSub (DVD) image subtitles and extract their text using external Vision Language Models.