1,503
▲ 13% since Mar 2024
Arena score of Anthropic's top model in March 2026, the highest of eight developers; xAI 1,495, Google 1,494, OpenAI 1,481
20242026

What the chart shows

In March 2026 Anthropic’s top model led the Arena at 1,503, followed by xAI (1,495), Google (1,494) and OpenAI (1,481): four developers within 22 points. Alibaba (1,449), DeepSeek (1,424) and Mistral AI (1,416) were further back.

In May 2023 OpenAI’s top model scored 1,322 against Google’s 1,117. Meta’s best score has stayed between 1,332 and 1,336 since August 2024. In February 2025 DeepSeek came within 5 points of the leader, OpenAI (1,400 against 1,405).

Key numbers

MeasureValue
Anthropic, Mar 20261,503
xAI, Mar 20261,495
Google, Mar 20261,494
OpenAI, Mar 20261,481
Alibaba, Mar 20261,449
Meta, Mar 20261,335

How to read it

  • Each line is the highest score of any model from that developer in each month (Arena historical leaderboard, public, style control on, exported March 2026). Lines start when a developer’s first model was ranked.
  • The report says DeepSeek briefly matched and surpassed the top US systems in early 2025; in the data its best month (February 2025) is 5 points below OpenAI, so we do not repeat that claim.
  • Arena scores have no true zero, so the vertical axis starts at 900 to show the gaps; compare differences between lines, not their heights above the axis.
Source: Arena, 2026, via the AI Index Report 2026 (Stanford HAI), figure 2.1.4: Performance of top models on the Arena by select providers. Chart and text by AI Adoption.
Numbers checked against the published data on 20 October 2026.
Download the data (CSV) · Cite this chart