Frontier Log

Meituan

LongCat-Flash-Chat

Announced on meituan.com

Benchmark results

BenchmarkScoreSource check
SWE-bench Verifiednon-thinking60.4%
CorroboratedThe same number is printed in 1 other lab document: 1
AIME 2024No tools · average of 10 runs; non-thinking70.42%
CorroboratedThe same number is printed in 1 other lab document: 1
AIME 2025No tools · average of 10 runs; non-thinking61.25%
Terminal-Bench 1.0non-thinking39.51%
CorroboratedThe same number is printed in 1 other lab document: 1
MMLU-Pronon-thinking82.7%
CorroboratedThe same number is printed in 2 other lab documents: 1 2
LiveCodeBenchnon-thinking; problem window not stated48.02%
CorroboratedThe same number is printed in 1 other lab document: 1
τ²-benchTelecom · average of 4 runs; non-thinking73.68%
τ²-benchRetail · average of 4 runs; non-thinking71.27%
τ²-benchAirline · average of 4 runs; non-thinking58.00%
Sources

Reported by Meituan

Related events

All models