Released: 2026-07-20
Key:
Score Threshold:
google/gemini-3-flash-preview
Score: 4954
Distance: 18km
Lat: 19.349666
Lng: -98.98663
The architecture shows typical Mexican urban construction with exposed rebar on flat roofs, common for future expansion. The vegetation, specifically the cypress and street-side dense rounded trees, is very common in the Valley of Mexico. The sun position and lighting suggest a location in the northern hemisphere, likely the suburbs of Mexico City such as Chimalhuacán or Iztapalapa.
anthropic/claude-haiku-4.5
anthropic
Score: 3379
Distance: 783km
Lat: 25.2865
Lng: -103.4045
This image shows characteristic features of a Mexican residential area. The architecture includes a multi-story brick/concrete building with flat roof typical of Mexican construction, metal security gates and fencing common in Mexican neighborhoods, Mediterranean-style cypress trees, and the overall urban layout and infrastructure. The style of street furniture, the condition of the pavement, and the general aesthetic strongly suggest this is from northern Mexico, likely in Coahuila state based on the architectural patterns and urban design. The residential compound style with high walls and security measures is typical of this region.
anthropic/claude-sonnet-4.6
anthropic
Score: 4988
Distance: 5km
Lat: 19.4978
Lng: -99
The image shows a typical Mexican urban residential neighborhood with characteristic features: unfinished brick/concrete block construction, water tanks on rooftops, metal security gates, and the overall architectural style common to the Estado de México or outer Mexico City metropolitan area. The street layout, trees, and building style are very typical of municipalities like Ecatepec, Nezahualcóyotl, or similar areas in the State of Mexico surrounding Mexico City.
meta-llama/llama-4-maverick
meta
Score: 4970
Distance: 12km
Lat: 19.4326
Lng: -99.1332
The image appears to be a residential area with a mix of modern and possibly older architectural styles, featuring tan brick buildings, wrought iron gates, and a variety of trees. The presence of a car and the overall layout suggest a suburban or urban neighborhood. The style of the buildings and the presence of certain features like the tan brick and the types of trees are consistent with areas found in Mexico City. The specific latitude and longitude are a guess based on the similarity of the image to areas in Mexico City, particularly in the context of the architectural style and urban planning.
openai/gpt-5-image-mini
openai
Score: 4972
Distance: 11km
Lat: 19.41
Lng: -99.08
Suburban Mexican appearance: exposed brick block construction, flat-roofed 2–3 storey building with roof antennas, metal driveway gate and low concrete walls common in Estado de México/Greater Mexico City neighborhoods. The closely pruned round street trees and a tall narrow cypress are also typical of residential streets in the Mexico City metro area (e.g., Nezahualcóyotl/late‑20th century suburban developments east of central CDMX). Lighting, car types and pavement details match that region, so I place the guess in the eastern Mexico City metro (approx. Nezahualcóyotl).
moonshotai/kimi-k2.5
moonshot
Score: 4956
Distance: 18km
Lat: 19.345
Lng: -99.066
This appears to be a residential street in Mexico City based on several visual cues: the architectural style of the unpainted concrete block construction with exposed rebar on the roof (common in Mexican residential construction), the street layout with trees planted along the sidewalk, the style of metal gates, and the Google Street View watermark style. The vegetation and climate suggest a high-altitude subtropical location. The concrete construction, unpainted buildings, and specific urban layout are characteristic of working-class neighborhoods in Mexico City, particularly in areas like Iztapalapa, Gustavo A. Madero, or similar districts. The pine tree visible suggests this could be in the eastern or southern parts of the city where此类 vegetation is more common.
Each model receives the same Google Street View image. Models are tasked with guessing the location of the image using only visual clues from the street view. Their answers are evaluated without feedback or retries.