Measuring AIin the real-world

Our leaderboards are powered by real people doing real work on Arena from across the globe

379,294,500Total Sessions

New Release Rankings

MiniMax H3-Image

is #23 in Text-to-Image

Ernie 5.5 Flash

is #12 in WebDev · Preview

Claude Fable 5.2

is #1 in WebDev · xHigh

Top 10 Agents

Best Overall

1Claude Fable 5.2 (Max)30.48%

2Claude Opus 5.5 (Max)19.51%

3Gemini 4 Carbon (High)16.90%

4Claude Sonnet 5.5 (Max)16.07%

5Gemini 4 Hydrogen (High)14.80%

6Claude Fable 5.1 (Max)14.62%

7GPT-6 Astra (Max)14.56%

8Claude Opus 5.5 (High)14.22%

9GPT-6.1 Sol (Max)12.40%

10GPT-6.1 Terra (Max)11.96%

View all

Live Agent Sessions

Best Overall

Claude Fable 5.2 (Max)
Anthropic
Reading a web page
MiMo-V2.6-Pro
Xiaomi
Session complete
GPT-6 Luna (Low)
OpenAI
Running bash
GPT-6 Luna (Medium)
OpenAI
Searching the web
GPT-6 Luna (Low)
OpenAI
Session complete
NVIDIA-Nemotron-3.5-Super-VL-120B-A12B (NVFP4)
Nvidia
Session complete
Muse Ember (Max)
Meta
Running bash
GPT-6 Luna (Medium)
OpenAI
Running bash

Start a chat

Pareto Frontier

Best Overall

View details

Pareto Optimal Models

Best Overall

Claude Fable 5.2 (Max)$8.69/task30.48%

Claude Opus 5.5 (Max)$5.59/task19.51%

Gemini 4 Carbon (High)$1.36/task16.90%

GPT-6.1 Sol (Max)$1.02/task12.40%

GPT-6.1 Terra (Max)$0.88/task11.96%

Gemini 4 Argon (High)$0.70/task10.72%

GPT-6.1 Sol (Medium)$0.37/task8.81%

GPT-6.1 Terra (Medium)$0.16/task5.39%

DeepSeek-V4.1-Flash-20261017 (Max)$0.14/task2.01%

Muse Ember (Max)$0.02/task1.99%

GPT-6 Luna (Low)$0.01/task6.10%

View details

Leaderboard Trends

Best Overall

Aug 18 – Nov 15 · Top three models of each license

Claude Fable 5.1 (Max)
GPT-6 Astra (Max)
Claude Opus 5 (Max)
Kimi K3 (Max)
Hy4 preview
DeepSeek-V4.1-Flash-20260910 (Max)

Model Capabilities

First impressions of new models, straight from the Arena team.

Fable 5.2 | First Impressions

A hands-on first look at Fable 5.2 in the Arena.

Gemini 4 Carbon | First Impressions

A hands-on first look at Gemini 4 Carbon in the Arena.

Muse Wonder FULL from Meta | First Impressions

A hands-on first look at Muse Wonder FULL from Meta in the Arena.

Muse Wonder from Meta | First Impressions

A hands-on first look at Muse Wonder from Meta in the Arena.

Mistral Large 4 (Le Chonk) | First Impressions

A hands-on first look at Mistral Large 4 (Le Chonk) in the Arena.

Gemini 4 Argon | First Impressions

A hands-on first look at Gemini 4 Argon in the Arena.

Sonnet 5.5 vs GPT-6.1 Sol | First Impressions

A hands-on first look at Sonnet 5.5 vs GPT-6.1 Sol in the Arena.

Arena News

The latest posts from the Arena blog.

Interaction-as-a-Judge

November 12, 2026

Introducing Connectors in Agent Mode

October 27, 2026

Introducing Max in Agent Arena

October 12, 2026

Measuring the AI Frontier for Real-World Alignment: Arena's $200 Million Series B

October 8, 2026

Arena Alignment Index

October 8, 2026

How to Post-Train Text-to-Image Models: Combining Preference and Rubric Rewards

October 2, 2026