Skip to content

The Aggregate

Unified LLM rankings, daily

Models What's New Trends Benchmarks Skill Maps
How It Works Guesswork Metrics
The Aggregate
Loading data...

Tau3 Banking — leaderboard

Metric: Success Rate (%). Source: artificialanalysis.ai. 168 models tracked.

Top models

#ModelScore
1Kimi K333.4
2GPT-5.6 Sol (Max)32.99
3GPT-5.6 Sol (xHigh)32.65
4Grok 4.5 (High)32.58
5GPT-5.6 Terra (Max)31.75
6GPT-5.5 (xHigh)31.34
7GPT-5.6 Sol (High)30.58
8Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort)30.52
9GPT-5.4 (xHigh)30.31
10GPT-5.5 (High)29.48
11Claude Opus 4.7 (Adaptive Reasoning, Max Effort)28.87
12Claude Sonnet 5 (Adaptive Reasoning, Max Effort)28.25
13Claude Opus 4.8 (Adaptive Reasoning, Max Effort)27.63
14GPT-5.6 Luna (Max)27.22
15GLM-5.2 (Max)26.8

Interactive version: theaggregate.ai/benchmark?slug=tau3-banking · How the rankings work · Data refreshed daily, snapshot 2026-07-22.

Built by Mikhail Doroshenko — AI researcher, co-author of Humanity’s Last Exam. Independent project; no affiliation with any AI lab.