Text2KGBench-LettrIA: F1 Scores Over Time (Generalization Benchmark)
Metric:
Overall F1
Entities (E)
Attributes (A)
Properties (P)
Relations (R)
Show:
All Models
Closed/Proprietary Only
Open-Weights Only
Benchmark:
All Benchmarks
Text2KGBench (2023)
Text2KGBench-LettrIA (2025)
Anthropic
Google
OpenAI
Mistral
Alibaba
Microsoft
Meta
Other
●
Closed/Proprietary
▲
Open-Weights
Note: Text2KGBench (2023) scores use DBpedia-WebNLG dataset. Text2KGBench-LettrIA (2025) uses refined DBpedia-WebNLG with improved annotations.