Live data from Hacker News

Wandr Benchmark: Evaluating Research Agents That Must Search Wide and Deep

research.perplexity.ai

Post reply on HN

Print view View original

1–2 of 2 posts

Wandr Benchmark: Evaluating Research Agents That Must Search Wide and Deep

#1

Post by tagawa » Fri, Jul 17, 2026, 12:45 AM UTC

Wandr Benchmark: Evaluating Research Agents That Must Search Wide and Deep
research.perplexity.ai

Re: Wandr Benchmark: Evaluating Research Agents That Must Search Wide and Deep

#2

Post by tagawa » Fri, Jul 17, 2026, 12:46 AM UTC

Repo with benchmark tasks, evaluation harness, tech report:

https://github.com/perplexityai/wandr

Post reply on HN