skill-models-dev

A Claude Code skill for querying AI model specifications, pricing, and capabilities from the models.dev open-source database.

Features

• Search models by name or ID across all providers
• Compare models side-by-side (context window, cost, capabilities)
• List providers and their available models
• Get detailed specs including pricing, limits, and capabilities

Installation

Option 1: Git Clone (Recommended)

cd ~/.claude/skills
git clone https://github.com/webup/skill-models-dev.git models-dev

Option 2: Manual Download

1. Download or copy the folder
2. Rename to models-dev and place in your skills directory:
   • macOS/Linux: ~/.claude/skills/models-dev/
   • Windows: %USERPROFILE%\.claude\skills\models-dev\

Directory Structure

~/.claude/skills/models-dev/
├── SKILL.md              # Skill definition (required)
├── README.md             # This file
└── scripts/
    └── fetch_models.py   # Query script

Usage

The skill activates automatically when you ask Claude Code questions like:

• "What's the context window for GPT-4o?"
• "Compare Claude vs GPT"
• "How much does Gemini Pro cost?"
• "List all OpenAI models"
• "Which models support reasoning?"

Manual Commands

You can also run the script directly:

# List all providers
python ~/.claude/skills/models-dev/scripts/fetch_models.py --providers

# Search models
python ~/.claude/skills/models-dev/scripts/fetch_models.py --search claude

# Get specific model info
python ~/.claude/skills/models-dev/scripts/fetch_models.py gpt-4o

# List models from a provider
python ~/.claude/skills/models-dev/scripts/fetch_models.py --provider anthropic

# Compare two models
python ~/.claude/skills/models-dev/scripts/fetch_models.py --compare claude-3-7-sonnet-20250219 gpt-4o

Example Output

Search Results

## Claude Sonnet 3.7
**Model ID:** `claude-3-7-sonnet-20250219`
**Provider:** Anthropic
**SDK Package:** `@ai-sdk/anthropic`
**Capabilities:** Reasoning, Tool Calling, Attachments, Temperature Control
**Modalities:** Input: text, image, pdf | Output: text
**Limits:** Context: 200,000 | Max Output: 64,000
**Cost:** Input: $3/M tokens | Output: $15/M tokens
**Knowledge Cutoff:** 2024-10-31
**Release Date:** 2025-02-19

Model Comparison

| Attribute | Claude Sonnet 3.7 | GPT-4o |
|-----------|-------------------|--------|
| Provider | Anthropic | OpenAI |
| Reasoning | Yes | No |
| Tool Calling | Yes | Yes |
| Context Window | 200,000 | 128,000 |
| Max Output | 64,000 | 16,384 |
| Input Cost | $3/M | $2.5/M |
| Output Cost | $15/M | $10/M |

Available Data

Each model includes (when available):

Field	Description
Model ID	Identifier for AI SDK integration
Provider	Company/platform (Anthropic, OpenAI, Google, etc.)
SDK Package	npm package for AI SDK
Capabilities	Reasoning, Tool Calling, Structured Output, Attachments
Modalities	Input/output types (text, image, audio, video, pdf)
Limits	Context window, max input/output tokens
Cost	Per-million token pricing (input, output, cache)
Dates	Knowledge cutoff, release date, last updated
Status	alpha, beta, deprecated
Open Weights	Whether weights are publicly available

Data Source

Data is fetched from models.dev API - a community-maintained open-source database of AI model specifications.

Requirements

• Python 3.7+
• No external dependencies (uses stdlib only)

License

MIT License

Contributing

Contributions welcome! Please submit issues and pull requests to improve the skill.

---

Created for use with Claude Code