Azure tools and practical guides. Feedback on LinkedIn.

← All tools

AI Leaderboard

Who leads the large language model race right now? Rankings come from the LMArena leaderboard, built from millions of blind, side-by-side human votes, enriched with release dates, context windows and API prices from OpenRouter.

Last refreshed: Loading... Arena snapshot: — Updated weekly by a GitHub Actions workflow

Highlights from the overall text leaderboard (style control on).

—
#1 model

 

—
Ranked models

Text arena

—
Organizations

With a ranked model

—
Open in top 20

 

—
Votes counted

Across ranked text models

—
Newest release

 

Ranking

Arena scores are Elo-style ratings: a 100-point gap means the higher model wins about 64% of head-to-head votes. Models whose confidence intervals overlap can share a rank.

Filter the ranking

Loading the leaderboard…

LLM ranking
RankTrendModelArena scoreVotesLicensePrice in / out
$ per 1M tokens
Context

Latest model releases

The newest models added to the OpenRouter catalogue, with their Arena text rank when they have already been rated.

Latest LLM releases
AddedModelArena rankContextPrice in / out
$ per 1M tokens
Capabilities

Top organizations

Labs ranked by their best model on the overall text leaderboard, with how many of their models reach the top 10 and top 50.

Sources and methodology