Yandex has announced a new update for Yandex Maps that expands how walking routes are suggested. Instead of focusing mainly on the fastest way to reach a destination, the service can now recommend routes that match different situations, preferences, and accessibility needs.
The company said that walking is often influenced by context. A route that works well on a busy weekday may not be the preferred choice during a weekend stroll or while exploring a new city.
New Route Characteristics
According to Yandex, walking routes in Yandex Maps can now include clear characteristics that help users choose a path that fits their needs.
Examples include:
- Leisurely
- Lively
- Easier
- With ramps
- Fewer stairs
- No stairs
- Optimal
- Fast
The company noted that alternative routes previously often appeared as variations of the same path with different travel times. The new system is designed to provide routes that reflect different walking experiences rather than simply longer or shorter versions of the same route.
Routes Designed For Different Moments
Yandex shared several examples of how the new route options can be used.
For weekend walks, the Maps app may suggest a Leisurely route that passes through parks, waterfront areas, and local landmarks where possible.
For people exploring an unfamiliar neighborhood after dark, a Lively route may prioritize busier streets and avoid quieter sections when possible. The goal is to provide route suggestions that better align with the situation and the user’s preferences.
The company said the best route is not always the fastest one. Instead, the most suitable route can change depending on the time, location, and purpose of the walk.
AI And Mapping Data Behind The System
Yandex said that artificial intelligence plays a major role in generating these route recommendations.
Behind the scenes, AI analyzes a wide range of signals related to the urban environment. These include parks, public spaces, street types, elevation changes, stairs, and local activity patterns.
The company explained that one of the biggest challenges was teaching the system to understand concepts that people naturally recognize but are difficult for machines to define.
Yandex highlighted examples such as determining what makes a route feel “pleasant” or “lively.” Because there is no simple rule for these qualities, the company combined large language models (LLMs), machine learning, and mapping data to help train the system.
Focus On Real-World Preferences
According to Yandex, the result is navigation that better reflects how people experience routes in the real world rather than focusing only on travel time.
The company said the update is intended to help users choose paths that match their current needs, whether they are looking for a scenic walk, a more active area, or an accessible route with fewer obstacles.
Yandex summarized the idea behind the feature with a simple message: the fastest route is not always the best route, and sometimes the best route is the one that fits the moment.
Sources
- Yandex (@yandexcom) posts on X, May 29, 2026.
