VAKRA Benchmark

What is VAKRA?

Evaluates multi-hop tool-use agents across live executable APIs and retrieval, scoring via policy adherence, exact match, and groundedness judges.

Released

2026-08-12

Evaluates

General AI, Information retrieval, Language & Knowledge, Reasoning

Openness

unknown

Importer

Claire Radar

Review status

ai-reviewed

Reported scores

0

Source provenance

• Original evidence https://arxiv.org/abs/2608.12282

VAKRA paper, code and dataset

• Paper https://arxiv.org/abs/2608.12282
• Code https://github.com/IBM/VAKRA
• Dataset https://huggingface.co/datasets/ibm-research/VAKRA

Which sources cite VAKRA?

• VAKRA

Related General AI, Information retrieval, Language & Knowledge, Reasoning benchmarks

• $\tau^\tau$-Bench: An Environment for End-To-End, Realistic Agent Construction
• 3DCodeBench
• 5-Dialects-BN
• A Large-Scale Dataset and Benchmark
• A-OKVQA
• A3S-Bench
• AARRI-Bench
• ABC-Bench