Released: 2026-02-26
Key:
Score Threshold:
google/gemini-3-flash-preview
Score: 4142
Distance: 377km
Lat: 18.397071
Lng: 105.151745
The vegetation consisting of ferns and dense tropical shrubbery on a steep embankment, combined with the narrow, weathered asphalt road and mountainous backdrop, is characteristic of the rural highland regions of Southeast Asia, specifically Central Vietnam or Laos. The camera quality and specific road style are common in Vietnamese mountainous provinces like Ha Tinh or Nghe An.
anthropic/claude-haiku-4.5
anthropic
Score: 3172
Distance: 910km
Lat: 13.7567
Lng: 100.5938
This image shows a rural paved road with dense tropical vegetation on both sides, including vines and ivy-covered walls. The landscape features lush green vegetation typical of Southeast Asia, with mountains visible in the distance. The road quality, vegetation type, and terrain characteristics are consistent with rural areas in Thailand, particularly in northern or northeastern regions. The dense ivy coverage on the right side and the tropical climate indicators suggest this is likely from Thailand’s countryside, possibly in the Chiang Mai or Nan province areas where such rural roads with steep vegetation-covered embankments are common.
meta-llama/llama-4-maverick
meta
Score: 2170
Distance: 1670km
Lat: 27.0429
Lng: 88.2653
The image shows a winding road surrounded by dense greenery and a hill on the right side. The road appears to be in a rural or mountainous area. The presence of lush vegetation and the style of the road suggest a location in a region with high rainfall and dense forests, possibly in the northeastern part of India or nearby regions. The specific location could be in Sikkim, India, given the terrain and vegetation.
anthropic/claude-sonnet-4.6
anthropic
Score: 77
Distance: 8348km
Lat: 44.2
Lng: 13.8
The image shows a narrow rural road with dense green vegetation on the right side (a steep embankment covered in lush plants) and trees on the left, with hills visible in the background. The road surface, vegetation type, and landscape are consistent with the Balkans or northern Italy/Adriatic coastal region. The lush subtropical-looking vegetation on the hillside and the road style suggest this could be in Croatia or Bosnia-Herzegovina along the Adriatic hinterland, or possibly northern Italy near the Adriatic coast. The overall appearance suggests the Dinaric Alps region or the Italian/Slovenian/Croatian border area. I’ll estimate a location in the Croatian coastal hinterland area.
x-ai/grok-4-fast
x-ai
Score: 1063
Distance: 3097km
Lat: 10.0881
Lng: 77.0593
The image depicts a narrow, winding paved road through dense, lush green vegetation on steep hillsides, with misty weather, which is characteristic of the tropical highland regions in Kerala, India, particularly around Munnar known for its tea estates and forested hills. The fern-like plants and broadleaf trees match the subtropical climate there.
openai/gpt-5-image-mini
openai
Score: 2865
Distance: 1114km
Lat: 22.3805
Lng: 114.3142
Narrow single-lane concrete mountain road with dense subtropical vegetation and abundant ferns on the cut slope — typical of Hong Kong country parks (Sai Kung / New Territories). Pavement texture and roadside drainage resemble Hong Kong rural roads rather than mainland China or SE Asia asphalt roads, so I place this in the Sai Kung/New Territories highlands (approx.).
moonshotai/kimi-k2.5
moonshot
Score: 1954
Distance: 1879km
Lat: 16.4167
Lng: 120.6333
This image shows a narrow paved road winding through dense, lush tropical vegetation with a steep embankment on one side and forest on the other. The road condition (narrow, paved but unmarked), the type of vegetation (dense ferns, broadleaf tropical plants, bamboo), and the mountainous terrain are characteristic of rural roads in the Cordillera region of Luzon, Philippines, particularly near areas like Sagada or Baguio. The tropical highland climate produces exactly this type of dense greenery and road infrastructure seen in the image.
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.