LLMs.txt Shows No Measurable Impact on AI Citations

A recent study by SE Ranking, encompassing an extensive analysis of approximately 300,000 domains, reveals that the llms.txt file currently shows no measurable impact on how frequently a website is cited by major Large Language Models (LLMs). This finding suggests that, despite discussions around its potential, the llms.txt standard is not yet delivering tangible benefits for AI citation frequency or overall AI visibility.

What the Data Reveals

Low Adoption Rates

The analysis uncovered remarkably low adoption rates for llms.txt. SE Ranking's crawl identified the file on just 10.13% of the domains surveyed, meaning nearly nine out of ten websites have not implemented it. This sparse usage is significant, especially given that llms.txt has sometimes been discussed as an emerging standard for controlling AI visibility. The data, however, points to scattered experimentation rather than widespread adoption, with usage fairly consistent across different traffic tiers and not concentrated among high-profile brands. Interestingly, high-traffic sites showed a slightly lower propensity to use the file compared to mid-tier websites in the dataset.

No Measurable Link to LLM Citations

To rigorously evaluate the llms.txt file's influence on AI visibility, SE Ranking meticulously analyzed domain-level citation frequency within responses generated by leading LLMs. Their methodology included statistical correlation tests and an XGBoost model to ascertain each factor's contribution to citations. A key revelation was that removing the llms.txt feature from their model actually improved its accuracy. This led SE Ranking to conclude that llms.txt "doesn’t seem to directly impact AI citation frequency. At least not yet." Furthermore, simpler statistical methods also yielded no significant correlation between the presence of the file and citation outcomes.

Alignment with Platform Guidance

SE Ranking's findings largely align with official guidance from major AI platforms, though precision is key. Google, for instance, has not indicated that llms.txt serves as a signal for its AI Overviews or AI Mode. In its AI search guidance, Google emphasizes that its AI search capabilities evolve from existing Search systems and signals, without any mention of llms.txt as an input.

Similarly, OpenAI's crawler documentation primarily focuses on robots.txt controls. While OpenAI advises allowing OAI-SearchBot via robots.txt for discovery, it does not state that llms.txt influences ranking or citation frequency. Although some SEO logs occasionally show GPTBot fetching llms.txt files, SE Ranking clarifies that this occurs infrequently and does not appear to be linked to citation outcomes. This collective evidence suggests that even if certain AI models access the file, it is not currently influencing citation behavior on a broad scale.

What This Means for You

For webmasters and SEO professionals, the implications are clear. While implementing llms.txt remains a low-risk, straightforward step for potential future adoption, it is crucial to manage expectations. If the objective is a near-term boost in AI visibility or citations within LLM answers, the current data indicates that such an outcome is unlikely. Consequently, llms.txt should be viewed as an experimental, early-stage tactic for AI visibility, suitable for testing if it aligns with existing workflows, but not yet a proven or reliable lever for immediate impact.