LLM Model Reference Hub

A comprehensive, fast, and clean web application serving as the definitive reference for current top LLM API models.

Features

Core Experience

• Model Directory: Clean, responsive grid interface showing major LLM providers (OpenAI, Anthropic, Google, Mistral, etc.).
• Detailed Model Cards: Displays essential info like pricing (input/output per 1M tokens), context window, and key capabilities.
• Code Snippets: Ready-to-use, copy-pasteable snippets for Python, JavaScript, and cURL for every model.
• llms.txt: Machine-readable endpoint at /llms.txt following the llmstxt.org standard for AI agent consumption.

Advanced Filtering & Search

• Real-time Search: Instantly filter models by name, provider, or API string.
• Capability Filtering: Clickable chips to filter models by specific capabilities (e.g., "Multimodal", "Coding", "Reasoning").
• Multi-Chip Support: Select multiple capabilities to find models that match all criteria.
• URL Persistence: All filters (search, provider, capabilities) are synced to the URL, making views easily shareable.

Technical

• Dark Mode: Fully responsive design with automatic dark mode support.
• Static Optimization: Built with Next.js App Router for maximum performance.

Getting Started

Prerequisites

• Node.js 18+
• npm

Installation

1. Clone the repository:

   git clone <repository-url>
   cd llm-reference-hub

2. Install dependencies:

   npm install

3. Run the development server:

   npm run dev

4. Open http://localhost:3000 with your browser.

Path to Full Automation

The project currently includes a skeleton script (scripts/update_models.ts) for automated updates. To achieve full automation, the following steps are required:

1. Search API Integration

• Missing: Integration with a web search API (e.g., Tavily, Brave Search, or Perplexity).
• Goal: The script needs to actively search for queries like "OpenAI GPT-4o pricing", "Anthropic Claude 3.5 release date", etc.

2. Content Parsing & Extraction

• Missing: Logic to parse search results or scrape official documentation pages.
• Goal: Extract structured data (pricing, context window, model names) from unstructured web content.

3. Verification & Diffing

• Missing: Logic to compare fetched data against the existing data/models.json.
• Goal: Identify discrepancies (e.g., price drops, new models) and generate a diff.

4. Alerting / PR Creation

• Missing: Integration with GitHub API or a notification service.
• Goal: Automatically create a Pull Request with the updated JSON data or send an alert (Slack/Discord) for human review when changes are detected.

Current Automation Status

The scripts/update_models.ts file currently loads the local data and iterates through it, serving as a placeholder for the logic described above.

To run the current script:

npx tsx scripts/update_models.ts

Adding Models Manually

To manually add a model, edit data/models.json. The schema is defined in lib/types.ts.

Deployment

This project is optimized for deployment on Vercel.

1. Push your code to GitHub.
2. Import the project in Vercel.
3. Deploy.

License

MIT