It’s that time of year again when college students around the country get a week to let loose. Many will seek out beaches and nightlife, and undoubtedly AI models will help many students find their Spring Break destinations. Where will these models send them?
We asked six large-language models (LLMs) to recommend Spring Break destinations for U.S. college students in four categories: best U.S. destinations, best European destinations, best international beaches and best budget-friendly destinations. We used prompts and categories U.S. college students frequently ask, taking care not to introduce bias into our prompts by using location or region-specific terms.
We found broad agreement across the models about where students should travel (Mexico, Miami and Prague, to name a few). We also observed that most models rarely cautioned would-be travelers about recent upheaval in Mexico, even when recommending locations named in February headlines about cartel violence and stranded travelers.
Florida & Mexico Dominate Suggestions
Mexico (and Cancun more specifically) appeared in the majority of responses across models, as did Florida locales like Miami, Panama City Beach and Key West. In the non-European categories, Miami or Mexico were among the five most-mentioned locations for all six models.
The distribution of the top 10 destinations looked fairly consistent across models. The most frequently mentioned destination often appeared in 70-85% of responses, and the 10th most-mentioned destination often appeared in 20-35% of responses, regardless of model or category.
Beyond the top 10, however, the models' recommendations largely diverge, a reality of LLMs that transcends topic as each model pulls from different sources to different degrees, have different foundational knowledge and behave differently in other ways.
There were also noticeable differences in the frequencies of mentions for any one destination. For example, the percentage of responses that named Mexico as a top destination for international beach seekers ranged from 65% on Perplexity to 88% on Google AI Mode.
See the top 10 recommendations for each model and category below:
Mexico and cities therein were frequently mentioned destinations for beach seekers and budget travelers. Mexican locales are apparently so irresistible that Perplexity couldn’t help but include Cancun as one of the best U.S. destinations (the only appearance of a non-U.S. city in the recommendations for U.S. destinations).
That’s not surprising, given the historic popularity of many Mexican resorts and beaches, but would-be travelers to Mexico certainly have reason for pause in light of recent events.
Models Diverge on Safety Warnings
The Mexican army killed the leader of a powerful cartel during an attempt to capture him on February 22. In response, the cartel burned buildings and cars, blocking roads and causing chaos at Guadalajara’s airport. The upheaval spread throughout the region, including popular tourist destinations like Puerto Vallarta.
Tourists were told to shelter in place as airlines canceled flights. By February 25, the U.S. Department of State had reopened consulates and advised U.S. citizens to resume “standard levels of precaution,” and flights returned to normal.
None of that news showed up in Copilot’s responses when the LLM recommended travelers go to Guadalajara. Instead, the model simply touted Guadalajara’s advantages, like “low-cost food, hostels, and transportation” and that the “street food is both delicious and inexpensive, and intercity buses are budget-friendly.”
Gemini, by contrast, specifically mentioned the February unrest in some of its responses, including these:
- “While resort areas like Cancun and Cabo remain heavily patrolled, recent cartel-related unrest (February 2026) has led to temporary flight disruptions and road closures in certain states.”
- “As of March 2026, the U.S. has issued specific ‘Exercise Increased Caution’ warnings for Mexico due to recent cartel-related unrest (specifically after the events in February 2026).”
Copilot was not alone in omitting these types of warnings. Neither Perplexity nor Google AI Overview included them either.
For example, Google AI Overview in recommending Guadalajara said, “Guadalajara and Merida are highlighted for their walkability, culture, and affordable culinary scenes, while cities like Puerto Vallarta offer beach access without the premium prices of Cancun.” Perplexity similarly noted that Guadalajara “is highlighted as a low-cost, high-reward city with warm weather and affordable dining.”
However, no model recommended Guadalajara when asked explicitly, “What are the safest international destinations for college spring break?” Still, Copilot confidently recommended Puerto Vallarta in 23% of its responses to this question without any caveats, even noting at times that its recommendations were based on “recent travel safety reports,” “recent travel reporting” and “recent travel analyses.”
Perplexity included Puerto Vallarta in many more responses to this question (54%), but it consistently included warnings like the following:
- “Please note that safety can change, so always check current advisories and local conditions before booking.”
- “Keep in mind that safety can change, so check current government travel advisories and local conditions before booking.”
- “I can share generally accepted safe international spring break destinations, but please note safety conditions can change and you should verify current advisories before booking.”
Perplexity appears to include similar boilerplate cautionary statements in response to many questions, regardless if they ask about safety. The statements do not dive into detail about any particular destination included in its recommendations.
Copilot only occasionally included a disclosure, “Check the latest U.S. State Department advisories before booking,” as an item in a list of safety tips.
Methodology
At Evertune Research we track hundreds of brands across 250 categories. For this analysis, we ran custom prompts that mimic the types of prompts U.S. college students ask AI about Spring Break destinations. We ran our prompts on ChatGPT, Copilot, Gemini, Google AI Mode, Google AI Overview and Perplexity, repeating each prompt for statistically meaningful samples.
Evertune is the AI marketing platform for Generative Engine Optimization (GEO) that helps brands improve visibility in AI search by analyzing responses at scale and delivering actionable insights. Evertune works with leading brands across all verticals, including Finance, Retail and E-Commerce, Automotive, Pharma, Tech, Travel, Food and Beverage, Entertainment, CPG, and B2B. Founded by early team members of The Trade Desk, Evertune has raised $20M in funding from leading adtech and martech investors. Headquartered in New York City, the company has a growing team of more than 40 employees.