Model Details
Longcat Flash Thinking-2601
Quick answer: Longcat Flash Thinking-2601 is Meituan's updated thinking model, an iteration on Longcat Flash Thinking. It leads the Tau2 Airline agent benchmark at 76.5% success rate and scores 88.6% on Tau2 Retail. Proprietary.
At a Glance
Where Longcat Flash Thinking-2601 leads
- 76.5% Tau2 Airline — top score on dual-control airline agent tasks
- 88.6% Tau2 Retail — strong dual-control retail agent performance
Where it lags
- Proprietary: no open weights
- Fewer independently-verified benchmark results than its predecessor
Benchmark Leaderboards
This model isn’t on any benchmark leaderboard yet.