Skip to content

README.md

Latest commit

 

History

History
114 lines (85 loc) · 3.07 KB

README.md

File metadata and controls

114 lines (85 loc) · 3.07 KB

AI Model Selector Skill

A comprehensive guide to AI model selection and API implementation for AI coding agents. Updated January 2026.

What's Included

File Description
SKILL.md Main entry point with decision tree
models/openai.md GPT-5.2, GPT-5.1, GPT-4.1, o-series specs
models/anthropic.md Claude Opus/Sonnet/Haiku 4.5
models/google.md Gemini 2.5/3 Pro/Flash
models/deepseek.md DeepSeek V3.2, R1 (cheapest option)
models/embeddings.md Voyage, OpenAI, Gemini embeddings
use-cases/coding.md Best models for coding tasks
use-cases/reasoning.md Math, science, planning
use-cases/rag.md RAG pipeline guidance

Key Features

  • Decision tree for quick model selection
  • Exact model strings and API parameters
  • Critical gotchas (e.g., GPT-5.x doesn't support temperature)
  • Code snippets for each provider
  • Current pricing (January 2026)
  • Use-case specific configurations

Quick Reference (January 2026)

Flagship Models

Model Context Price (in/out per 1M)
GPT-5.2 400K $1.75 / $14
Claude Opus 4.5 200K $5 / $25
Claude Sonnet 4.5 200K (1M beta) $3 / $15
Gemini 2.5 Pro 1M $1.25 / $10

Budget Options

Model Context Price (in/out per 1M)
Claude Haiku 4.5 200K $1 / $5
Gemini 2.5 Flash 1M $0.30 / $2.50
DeepSeek-chat 128K $0.28 / $0.42

Installation

With npx skills (Vercel)

npx skills add jaymay549/ai-model-selector

With skild

skild install jaymay549/ai-model-selector

With Claude Code

/plugin marketplace add jaymay549/ai-model-selector

Manual

Copy the folder to:

  • Claude Code: ~/.claude/skills/ai-model-selector/
  • Cursor: .cursor/skills/ai-model-selector/
  • Codex: ~/.codex/skills/ai-model-selector/

Example Usage

When you ask your AI agent:

  • "What model should I use for coding?"
  • "How do I call the GPT-5.2 API?"
  • "Set up embeddings for RAG"

The agent will reference these files to give accurate, up-to-date answers.

Models Covered

Chat/Completion

  • OpenAI: GPT-5.2, GPT-5.2 Pro, GPT-5.2-Codex, GPT-5.1, GPT-5, GPT-4.1, o3, o4-mini
  • Anthropic: Claude Opus 4.5, Sonnet 4.5, Haiku 4.5
  • Google: Gemini 3 Pro, Gemini 2.5 Pro, Gemini 2.5 Flash
  • DeepSeek: V3.2 (deepseek-chat), R1 (deepseek-reasoner)

Embeddings

  • Voyage: voyage-3.5, voyage-3-large, voyage-code-3, voyage-law-2, voyage-finance-2
  • OpenAI: text-embedding-3-large, text-embedding-3-small
  • Google: gemini-embedding-001

Critical Gotcha

GPT-5.x and o-series models do NOT support temperature, top_p, or max_tokens. Use:

{
  "model": "gpt-5.2",
  "reasoning": { "effort": "medium" },
  "max_completion_tokens": 4096
}

Contributing

PRs welcome! Model specs change frequently - please include sources when updating.

License

MIT

Last Updated

January 28, 2026

Sources: Official documentation from OpenAI, Anthropic, Google, DeepSeek, Voyage AI