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

The best Work models,
ranked by real use.

Tracks everyday business tasks including contract review, structured data extraction, emails, spreadsheets, and planning docs.

378,706,203sessions rankedHow we rank →

Rising this week · Work

Claude Fable 5.1up 1.55 pts · #4 in WorkGemini 4 Hydrogenup 1.21 pts · #7 in WorkGemini 4 Carbonup 0.76 pts · #3 in Work

The podium · Work

Win share · this week
  1. Rank 1: Claude Fable 5.2Anthropic · 1542 score30.01%
  2. Rank 2: Claude Opus 5.5Anthropic · 1486 score18.60%
  3. Rank 3: Gemini 4 CarbonGoogle · 1470 score15.86%

Full ranking

Work leaderboard: win share, score, cost per task and votes
#ModelVotes7 days
1Claude Fable 5.2 (Max)Anthropic30.01%1542 ±3$8.02284,000
2Claude Opus 5.5 (Max)Anthropic18.60%1486 ±2$5.71251,733
3Gemini 4 Carbon (High)Google15.86%1470 ±1$1.42232,466
4Claude Fable 5.1 (Max)Anthropic15.23%1464 ±2$1.25208,199
5Claude Sonnet 5.5 (Max)Anthropic14.71%1459 ±2$1.95196,932
6GPT-6.1 Sol (Max)OpenAI12.60%1444 ±1$1.03174,665
7Gemini 4 Hydrogen (High)Google11.80%1438 ±2$0.92168,398
8GPT-6.1 Terra (Max)OpenAI10.90%1429 ±1$0.89154,131
9DeepSeek V4.1 FlashDeepSeek6.80%1396 ±1$0.12149,864
#130.01%
Score 1542 ±3$8.02/task284,000 votes
#218.60%
Score 1486 ±2$5.71/task251,733 votes
Score 1470 ±1$1.42/task232,466 votes
#415.23%
Score 1464 ±2$1.25/task208,199 votes
#514.71%
Score 1459 ±2$1.95/task196,932 votes
#612.60%
Score 1444 ±1$1.03/task174,665 votes
Score 1438 ±2$0.92/task168,398 votes
#810.90%
Score 1429 ±1$0.89/task154,131 votes
#96.80%
Score 1396 ±1$0.12/task149,864 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 1542 · 30.01% 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.

§ Work · Questions & explore

About the work board.

AI productivity toolsFunding newsAPI pricesAPI quickstart

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

Claude Fable 5.2 (Anthropic) leads work with a 30.01% real-work win share and a Bradley-Terry score of 1542, ahead of #2 Claude Opus 5.5 (18.60%).

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

DeepSeek V4.1 Flash balances #9 performance (score 1396) at $0.12 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 work 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.