Developing with LLMs and AI Tooling
Our documentation is now available in LLMs.txt format, optimized for AI reasoning engines. To better support platform-specific development, we've created separate documentation files for each platform:
llms-1.43-full-ios.txt
: Complete iOS SDK documentation (~120,000 tokens)llms-1.43-full-web.txt
: Complete Web SDK documentation (~500,000 tokens)llms-1.43-full-android.txt
: Complete Android SDK documentation (~120,000 tokens)
These documentation files are substantial in size, with token counts exceeding the context windows of many AI models. This guide explains how to download and effectively use these platform-specific documentation files with AI tools to accelerate your development process.
What are LLMs.txt files?#
LLMs.txt is an emerging standard for making documentation AI-friendly. Unlike traditional documentation formats, LLMs.txt:
- Presents content in a clean, markdown-based format
- Eliminates extraneous HTML, CSS, and JavaScript
- Optimizes content for AI context windows
- Provides a comprehensive view of documentation in a single file
By using our platform-specific LLMs.txt files, you'll ensure that AI tools have the most relevant and complete context for helping with your development tasks.
Downloading Platform-Specific Documentation#
Platform | Documentation Link |
---|---|
Platform iOS | Documentation Link https://cdn.img.ly/tooling/llms-1.43-full-ios.txt |
Platform Web | Documentation Link https://cdn.img.ly/tooling/llms-1.43-full-web.txt |
Platform Android | Documentation Link https://cdn.img.ly/tooling/llms-1.43-full-android.txt |
Handling Large Documentation Files#
Due to the size of our documentation files—particularly the Web documentation at ~500,000 tokens—most AI tools will face context window limitations. Standard models typically have context windows ranging from 8,000 to 200,000 tokens, making it challenging to process our complete documentation in a single session.
Recommended AI Model for Full Documentation#
For working with our complete documentation files, we recommend:
- Gemini 2.0 Flash: Available via Google AI Studio with a context window of 1-2 million tokens, capable of handling even our largest documentation file