BenchmarkSkillsBench

SkillsBench

SkillsBench is a benchmark for evaluating the skill level of models in using tools, focusing on the correctness of a single tool or API call, including argument construction, return value parsing, and error handling.

Timeline (last 90 days)

Oct 18

Ling-3.1-flash scored 68.70 on the SkillsBench benchmark.

Unverified50%

Oct 18

SkillsBench focuses on the correctness of single tool or API calls, including parameter construction, return value parsing, and error handling.

Unverified50%

All Facts (2)

Unverified

Ling-3.1-flash scored 68.70 on the SkillsBench benchmark.

50%Unverified

SkillsBench focuses on the correctness of single tool or API calls, including parameter construction, return value parsing, and error handling.

50%

Source Articles

Ling-3.1-flash Open-Sourced: 1673 on GDPval-AA, a New Choice for Office AgentsOct 18