# LLM Search News > How LLM search works: retrieval, crawlers, citations and the models behind them. News is compiled from primary sources (company announcements, official documentation, studies) and every story is dated. Each article is also available as clean Markdown at the URLs below. Full text of all articles: https://llmsearchnews.com/llms-full.txt ## Articles - [How often do AI search engines get answers and citations wrong?](https://llmsearchnews.com/ai-answer-accuracy-studies-2026.md): As of 17 November 2026, no single study establishes a universal error rate for AI search. (updated 2026-11-17) - [How much do AI crawlers crawl compared to what they send back?](https://llmsearchnews.com/ai-crawler-traffic-2026.md): Cloudflare Radar’s 17 November 2026 snapshot records 3,351.73 Perplexity crawls, 2,073.96 OpenAI crawls and 456.23 Anthropic crawls per observed referral over the last seven days. (updated 2026-11-17) - [What is query fan-out in AI search?](https://llmsearchnews.com/query-fan-out-explained-2026.md): Query fan-out turns one request into related searches across subtopics or sources; Google says AI Mode may run those searches concurrently. (updated 2026-11-17) - [Which model releases changed AI search in late 2026?](https://llmsearchnews.com/model-releases-search-november-2026.md): Google’s 12 November 2026 rollout of Gemini 4 Hydrogen into Google Search is the clearest documented change to consumer AI search in late 2026. (updated 2026-11-17) - [How does ChatGPT search decide which sources to cite?](https://llmsearchnews.com/how-chatgpt-search-chooses-sources.md): As of 17 November 2026, OpenAI documents ChatGPT Search as a multi-stage process that can rewrite prompts, query providers, review results, and issue follow-up searches. (updated 2026-11-17) ## Beats - [How it works](https://llmsearchnews.com/beats/how-it-works): Explainers on query rewriting, fan-out, retrieval and how answers pick their citations. - [Crawlers](https://llmsearchnews.com/beats/crawlers): GPTBot, ClaudeBot, OAI-SearchBot and friends: crawl volumes, referrals, robots.txt and blocking. - [Research](https://llmsearchnews.com/beats/research): Studies and papers on answer accuracy, citation faithfulness and how LLM search behaves. - [Models](https://llmsearchnews.com/beats/models): New GPT, Gemini, Claude and open-weight releases, and what each documents about search and browsing. ## About - [Editorial standards](https://llmsearchnews.com/methodology): sources, dating and corrections - [About](https://llmsearchnews.com/about): who publishes this site