Share your thoughts, 1 month free Claude Pro on usSee more

Auto-formalization on Prover-Bench

100Pass@8

LongCat-Flash-Prover

Updated 26d ago

Evaluation Results

Method	Links
LongCat-Flash-Prover 2026.03		100
LongCat-Flash-Prover 2026.03		95.2
Claude-Opus-4.5 2026.03		94.8
Goedel-V2-Formalizer-32B 2026.03		94.4
Gemini-3 Pro 2026.03		93
Goedel-V2-Formalizer-8B 2026.03		93
Kimi-K2.5 2026.03		91.7
ATF-32B 2026.03		89.6
DeepSeek-V3.2 2026.03		83
ATF-8B-Distilled 2026.03		83
StepFun-Formalizer-32B 2026.03		73.5
Kimina-Autoformalizer-7B 2026.03		67.4
StepFun-Formalizer-7B 2026.03		65.2