ChatGPT built its own search index "Labrador" — a family of 12+ vertical indexes, not just Bing
lilyraynyc · x · 2026-09-04
- Peec AI's Tomek Rudzki reveals ChatGPT runs its own retrieval index, internally called Labrador — overturning two years of SEO assumptions that it only used Bing/Google.
- Labrador is a family of vertical indexes: general web, PDF, YouTube, news (tiered by recency), arXiv, Wikipedia, local, finance, legal, medical, shopping, and images.
- Evidence: a resultsource field in server-side events (May–Jul 2026) only carried four values — Labrador plus three external scraping providers; Google antitrust testimony and job postings confirm a years-long deliberate build.
- ChatGPT is actively A/B testing its own index against scraped results, most visibly in shopping, and caches pages instead of fetching live.
- Bing/Google still matter: ChatGPT pulls from at least eight other providers, including Yelp, TripAdvisor, and Microsoft's Web IQ.
More from Infra
- KV caching: the fundamental optimization behind autoregressive LLM inference — alec_helbling · 2026-09-04
- WSJ: Data centers are a "lottery win" for workers in Richland Parish, Louisiana — robleclerc · 2026-09-04
- No orchestrator, no MCP: 4 peer agents on a gossip mesh research the web and deliver to Slack — Trainer_Intelligent · 2026-09-04
- DLSS 5 ships with NBA 2K27: costs half your 4K frame rate, worth it — ryanshrout · 2026-09-04
- Kastard: Edit ComfyUI Workflows Locally, Auto-Sync Models to RunPod GPUs — spacebearbug · 2026-09-04
- Running Qwen3.8 Flash NVFP4 on a single DGX Spark: 1M context at 37 tok/s — BLUECOW009 · 2026-09-04