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 Text models,
ranked by real use.

Captures long-form prose clarity, structured reasoning, factual synthesis, summarization, and multi-language instruction following.

378,706,203sessions rankedHow we rank →

Rising this week · Text

Claude Opus 5.5up 1.95 pts · #4 in TextGrok 4.7up 1.58 pts · #7 in TextMuse Wonder Fullup 0.95 pts · #3 in Text

The podium · Text

Win share · this week
  1. Rank 1: Gemini 4 CarbonGoogle · 1526 score22.40%
  2. Rank 2: Gemini 4 ArgonGoogle · 1522 score21.60%
  3. Rank 3: Muse Wonder FullMeta · 1511 score19.80%

Full ranking

Text leaderboard: win share, score, cost per task and votes
#ModelVotes7 days
1Gemini 4 Carbon (High)Google22.40%1526 ±10$1.42310,000
2Gemini 4 Argon (High)Google21.60%1522 ±5$0.70294,733
3Muse Wonder FullMeta19.80%1511 ±6$0.88241,466
4Claude Opus 5.5 (Max)Anthropic19.10%1507 ±6$5.71276,199
5Claude Fable 5.2 (Max)Anthropic18.75%1504 ±4$8.02265,932
6GPT-6.1 Sol (Max)OpenAI16.90%1495 ±7$1.03212,665
7Grok 4.7xAI15.40%1482 ±8$1.85188,398
8DeepSeek V4.1 FlashDeepSeek11.20%1446 ±7$0.12205,131
Score 1526 ±10$1.42/task310,000 votes
#221.60%
Score 1522 ±5$0.70/task294,733 votes
#319.80%
Score 1511 ±6$0.88/task241,466 votes
#419.10%
Score 1507 ±6$5.71/task276,199 votes
#518.75%
Score 1504 ±4$8.02/task265,932 votes
#616.90%
Score 1495 ±7$1.03/task212,665 votes
#715.40%
Score 1482 ±8$1.85/task188,398 votes
#811.20%
Score 1446 ±7$0.12/task205,131 votes

Pareto frontier

score vs $/task

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

Gemini 4 CarbonPareto optimal
Score 1526 · 22.40% win · $1.42/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.

§ Text · Questions & explore

About the text board.

AI research assistantsAnalysis newsAPI pricesAPI quickstart

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

Gemini 4 Carbon (Google) leads text with a 22.40% real-work win share and a Bradley-Terry score of 1526, ahead of #2 Gemini 4 Argon (21.60%).

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

DeepSeek V4.1 Flash balances #8 performance (score 1446) 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 text 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.