Can it handle large video files?

Yes, it includes patterns for Gemini's native video support as well as frame extraction strategies for models that only support static image inputs.

How can I reduce API costs for vision tasks?

The skill includes cost-optimization patterns such as using 'low' detail modes for simple classification and resolution resizing strategies to minimize token consumption.

Does it support object detection?

Yes, it provides specific logic for generating bounding boxes and performing region analysis, particularly optimized for Gemini 2.5's native detection capabilities.

Which models does this skill support?

The skill supports the latest multimodal models including GPT-5.2, Claude Opus 4.6, Gemini 2.5/3, and Grok 4, providing specific implementation code for each.

How does it handle PDF documents in Claude Code?

It utilizes Claude Code's advanced PDF handling features, supporting page-range parameters and incremental processing for documents exceeding 100 pages.

Vision Language Models

Name: Vision Language Models
Author: yonatangross

byyonatangross

•

データサイエンスとML

Integrates advanced multimodal capabilities from leading AI models for image analysis, document understanding, and visual reasoning.

Vision Language Models provides production-ready implementation patterns for integrating multimodal capabilities into Claude Code workflows. It enables developers to leverage state-of-the-art vision models—including GPT-5, Claude 4.5, and Gemini 2.5—to perform complex tasks such as automated document processing, visual question answering, multi-image comparison, and native video analysis. This skill is essential for building applications that require high-fidelity OCR, spatial reasoning, or the ability to 'see' and interpret visual data directly within a coding environment, complete with cost-optimization strategies and token management.

主な機能

01Advanced document and chart analysis with structured OCR patterns

0269 GitHub stars

03Cross-provider support for GPT-5, Claude 4.5, Gemini 2.5/3, and Grok 4

04Native video frame processing and multi-image reasoning capabilities

05Spatial awareness tools for bounding box and object detection

06Optimized token management and resolution-based cost efficiency patterns

ユースケース

01Building automated document extraction pipelines for complex PDFs, diagrams, and charts

02Developing visual comparison tools to track UI/UX changes and regression across image sequences

03Implementing visual QA and accessibility alt-text generation in modern web applications

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add yonatangross/orchestkit vision-language-models

For use in Claude.ai and ChatGPT

Download Skill