AI Model Rankings tells you which AI model to use. It combines head-to-head quality rankings voted on by millions of people with live prices and specs for more than 400 models, so you can see who is best, who is cheapest, and who gives the most quality per dollar.
Rankings cover text, coding and web development, vision, search, documents, agents, image generation and editing, and video. Text rankings break down further by task, such as coding, maths, creative writing, hard prompts, long queries and specific languages. Prices and specs come from a live catalogue, so new models appear the day they launch.
What you can do
- leaderboard — the current ranking for any arena and category, with rating, confidence and vote count
- best_value — models ranked by quality alongside price, with the best-value picks flagged
- search_models — filter the live catalogue by name, context length, input or output type, tool support and price
- compare_models — put two to four models side by side on price, context, features and ranking
Who it's for
Developers choosing a model for a product, teams trying to cut AI costs without losing quality, and anyone who wants an up-to-date answer to "which model is best for X?"
How to use it
- Start with leaderboard for the arena you care about, e.g. webdev for coding or text_to_image for images
- Use best_value to see which top models are worth their price
- Use search_models to find candidates that fit hard limits such as context length or a price ceiling
- Finish with compare_models on your shortlist
Getting started
Ask "what's the best model for coding right now?" or "cheapest model with a 1M context window". Rankings include their source credit; keep it when you share them.
