ENفا
Get API key
MODELSArena’s interaction judge tests what agents actually finish▼ 22%DeepSeek V4.1 Flash API price#1Claude Fable 5.2 · 30.48% win shareFUNDINGAnthropic targets $64.4B Nasdaq IPO as revenue surgesTOOLSCursor tops the tools directory this weekMODELSGoogle cuts Gemini 4 Carbon prices before Gemini 2.5 sunsetPOLICYFive frontier AI labs face coordinated SF safety lawsuits
Vigas Arena · updated just now

Measuring AI
in the real world.

Rankings built from real people doing real work on Vigas — every session, every blind vote, every dollar spent.

378,706,203sessions rankedHow we rank →

Rising this week · Overall

Claude Opus 5.5up 2.50 pts · #2 in OverallGemini 4 Carbonup 1.92 pts · #3 in OverallClaude Sonnet 5.5up 1.53 pts · #4 in Overall

The podium · Overall

Win share · this week
  1. Rank 1: Claude Fable 5.2Anthropic · 1548 score30.48%
  2. Rank 2: Claude Opus 5.5Anthropic · 1489 score19.51%
  3. Rank 3: Gemini 4 CarbonGoogle · 1476 score16.90%

Full ranking

Overall leaderboard: win share, score, cost per task and votes
#ModelVotes7 days
1Claude Fable 5.2 (Max)Anthropic30.48%1548 ±3$8.02498,000
2Claude Opus 5.5 (Max)Anthropic19.51%1489 ±2$5.71441,733
3Gemini 4 Carbon (High)Google16.90%1476 ±1$1.42418,466
4Claude Sonnet 5.5 (Max)Anthropic16.07%1470 ±3$1.95382,199
5Gemini 4 Hydrogen (High)Google14.80%1463 ±2$0.92312,932
6Claude Fable 5.1 (Max)Anthropic14.62%1458 ±2$1.25305,665
7GPT-6 Astra (Max)OpenAI14.56%1456 ±2$4.10334,398
8GPT-6.1 Sol (Max)OpenAI12.40%1446 ±1$1.03298,131
9GPT-6.1 Terra (Max)OpenAI11.96%1442 ±1$0.89279,864
10Gemini 4 Argon (High)Google10.72%1433 ±1$0.70284,597
11GPT-6 Luna (Low)OpenAI6.10%1418 ±1$0.15228,330
12MiMo-V2.6-ProXiaomi2.16%1394 ±1$0.16176,063
13DeepSeek V4.1 FlashDeepSeek2.01%1391 ±1$0.12214,796
14Muse Ember (Max)Meta1.99%1382 ±1$0.22148,529
#130.48%
Score 1548 ±3$8.02/task498,000 votes
#219.51%
Score 1489 ±2$5.71/task441,733 votes
Score 1476 ±1$1.42/task418,466 votes
#416.07%
Score 1470 ±3$1.95/task382,199 votes
Score 1463 ±2$0.92/task312,932 votes
#614.62%
Score 1458 ±2$1.25/task305,665 votes
#714.56%
Score 1456 ±2$4.10/task334,398 votes
#812.40%
Score 1446 ±1$1.03/task298,131 votes
#911.96%
Score 1442 ±1$0.89/task279,864 votes
#1010.72%
Score 1433 ±1$0.70/task284,597 votes
#116.10%
Score 1418 ±1$0.15/task228,330 votes
#122.16%
Score 1394 ±1$0.16/task176,063 votes
#132.01%
Score 1391 ±1$0.12/task214,796 votes
#141.99%
Score 1382 ±1$0.22/task148,529 votes

Pareto frontier

score vs $/task

Green models give the best score for their price. Tap any dot to inspect value vs. performance.

Claude Fable 5.2Pareto optimal
Score 1548 · 30.48% win · $8.02/task
API pricing

Live sessions

LIVE

No sessions in the last 24 hours yet. Send a request through the Vigas API.

Connect via API — start in 30s
§ Methodology

How the rankings are made.

Real sessions, not benchmarks

Win share counts how often a model finishes real tasks that users accept — across real API workloads, coding agents, and blind Arena tests.

Blind pairwise votes

In Arena, people compare two hidden models and vote. Scores use a Bradley–Terry model with 95% confidence intervals.

Cost is part of the score

Every ranking shows dollars per completed task, so you can see who is best for the money — not just best.

§ Overall · Questions & explore

About the overall board.

AI agent toolsModels newsAPI pricesAPI quickstart

What is the #1 AI model on the Overall leaderboard right now?

Claude Fable 5.2 (Anthropic) leads overall with a 30.48% real-work win share and a Bradley-Terry score of 1548, ahead of #2 Claude Opus 5.5 (19.51%).

Which top-10 model offers the lowest cost per task for overall?

Gemini 4 Argon balances #10 performance (score 1433) at $0.70 per completed task on Vigas.

How are Vigas leaderboard scores and win shares calculated?

Vigas combines real session outcomes across API workloads and agent harnesses with blind pairwise Arena votes using a Bradley-Terry statistical model with 95% confidence intervals.

Can I call the top overall models through one API?

Yes. Point any OpenAI-compatible client at https://api.vigas.ai/v1 with your Vigas API key or set model to vigas/auto to route across providers automatically.