AgentSkillsCN

models-dev

从models.dev数据库中查询AI模型的规格、定价与功能特性。当用户询问AI模型的参数(上下文窗口、Token限额、每Token的成本)、模型对比、服务商信息,或需要查询特定模型ID以集成AI SDK时,可使用此技能。例如,“GPT-4o的上下文窗口是多少?”“比较Claude与GPT”“Gemini Pro的价格是多少?”“列出OpenAI的模型”“哪些模型支持工具调用?”等查询均可触发此技能。

SKILL.md
--- frontmatter
name: models-dev
description: Query AI model specifications, pricing, and capabilities from models.dev database. Use when users ask about AI model parameters (context window, token limits, cost per token), model comparisons, provider information, or need to look up specific model IDs for AI SDK integration. Triggers on queries like "What's the context window for GPT-4o?", "Compare Claude vs GPT", "How much does Gemini Pro cost?", "List OpenAI models", or "What models support tool calling?".

Models.dev - AI Model Information Query

Query comprehensive AI model specifications from the models.dev open-source database.

Usage

Run the script at scripts/fetch_models.py:

bash
# List all providers
python scripts/fetch_models.py --providers

# Search models by name or ID
python scripts/fetch_models.py --search <query>

# Get specific model info
python scripts/fetch_models.py <model_id>

# List models from a provider
python scripts/fetch_models.py --provider <provider_name>

# Compare two models
python scripts/fetch_models.py --compare <model_id_1> <model_id_2>

Examples

bash
# Search for Claude models
python scripts/fetch_models.py --search claude

# Get GPT-4o details
python scripts/fetch_models.py gpt-4o

# List Anthropic models
python scripts/fetch_models.py --provider anthropic

# Compare Claude Sonnet 3.7 vs GPT-4o
python scripts/fetch_models.py --compare claude-3-7-sonnet-20250219 gpt-4o

Available Data

Each model includes (when available):

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

API Source

Data fetched from https://models.dev/api.json - a community-maintained open-source database.