# Perplexity — Q2D-Web: Evaluating First-Stage Retrievers at Scale

- Company: Perplexity (perplexity.ai)
- Announced: 2026-09-09
- Category: research-paper
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://www.perplexity.ai/hub/blog/q2d-web
- Record: https://forck.live/items/10750-q2d-web-evaluating-first-stage-retrievers-at-scale
- Subject: Perplexity

Perplexity introduced Q2D-Web, a private benchmark and public leaderboard for evaluating first-stage retrieval in agentic RAG systems. The benchmark includes 190 million web documents and 69,721 agent-reformulated queries across ten languages, with three relevance-judgment sets to reduce bias and false negatives.

## Evidence

Verbatim from https://www.perplexity.ai/hub/blog/q2d-web:

> Q2D-Web is built to evaluate embedding models on large-scale web search. It consists of 190 million web documents and 69,721 agent-reformulated queries in ten languages, sampled over nine months of PII-free production search traffic.

---

Record: https://forck.live/items/10750-q2d-web-evaluating-first-stage-retrievers-at-scale
Catalogue: https://forck.live/llms.txt
Current issue: https://forck.live/feed.md