Released: 2026-05-14
Key:
Score Threshold:
google/gemini-3-flash-preview
Score: 4709
Distance: 120km
Lat: 15.011684
Lng: 104.423985
The dry, reddish soil and flat agricultural landscape with bamboo and scattered tropical trees are characteristic of Northeast Thailand (Isan) during the dry season. The presence of simple wooden fencing and the specific vegetation patterns are common throughout provinces like Sisaket and Ubon Ratchathani.
meta-llama/llama-4-maverick
meta
Score: 4488
Distance: 216km
Lat: 14.410555
Lng: 103.121666
The image shows a dry field with a stream of water spraying from a pipe, surrounded by trees and a blue sky. The presence of a Google watermark and the style of the image suggest it is a Google Street View image. The dry conditions and the type of vegetation are consistent with regions that experience a dry season. The copyright date of 2023 indicates the image is relatively recent. Based on these observations and knowing that Google Street View covers many parts of the world, including rural areas, the location is likely in a region with a tropical savanna climate. The specific coordinates are guessed to be near Surin, Thailand, a region known for its agricultural fields and dry seasons.
anthropic/claude-haiku-4.5
anthropic
Score: 4719
Distance: 116km
Lat: 16.8661
Lng: 104.8655
This image shows a dry, flat landscape with sparse vegetation and scattered trees typical of Southeast Asia’s dry season. The terrain appears to be agricultural land or savanna-like environment. The vegetation pattern, soil color, and landscape characteristics are consistent with the Mekong region, likely in Cambodia or northeastern Thailand. The white line visible on the ground suggests a road or boundary marker. The flat terrain with scattered trees and dry conditions point to the Cambodian countryside, possibly in the central or northern regions during the dry season (November-May).
openai/gpt-5-image-mini
openai
Score: 4296
Distance: 303km
Lat: 13.352
Lng: 103.859
Flat, open dry field with reddish-brown soil, scattered bamboo and low tropical trees, and grazing cattle — typical of the dry-season countryside of northwest Cambodia (Siem Reap/Kampong Thom region). Vegetation mix and simple stick fences are common in rural Cambodia; sky and cloud type suggest tropical latitude. Overall landscape and land use point to rural Cambodia near Siem Reap, so I place the guess at 13.352, 103.859.
anthropic/claude-sonnet-4.6
anthropic
Score: 4316
Distance: 294km
Lat: 13.5
Lng: 103.5
The image shows a flat, dry agricultural field (likely a rice paddy in the dry season) with tropical vegetation in the background. The landscape, vegetation types, and general appearance are consistent with rural Cambodia or nearby Southeast Asian countries. The flat terrain with dry rice fields and mixed tropical trees suggests the Cambodian lowlands, possibly near Siem Reap or in the central plains of Cambodia.
x-ai/grok-4-fast
x-ai
Score: 4839
Distance: 65km
Lat: 16.1234
Lng: 103.5678
The image depicts a dry, open field with scattered bamboo and tropical trees, cows grazing, and a wire fence under a clear blue sky, characteristic of the dry season in northeastern Thailand. The vegetation, flat terrain, and rural pastoral setting match landscapes near Ubon Ratchathani province. The 2013 Google Street View timestamp aligns with coverage in this region.
moonshotai/kimi-k2.5
moonshot
Score: 4319
Distance: 293km
Lat: 13.45
Lng: 103.85
The image depicts a flat, rural landscape characteristic of the Cambodian countryside during the dry season. Key indicators include the tropical savanna vegetation with scattered deciduous trees, reddish-brown dry soil with seasonal puddles (suggesting a floodplain or rice paddy area), simple wooden fencing, and the overall flat terrain typical of the Tonle Sap basin or surrounding plains. The bright blue sky with scattered clouds and vegetation patterns are consistent with the Siem Reap or Kampong Thom provinces in Cambodia. The landscape matches Street View imagery from rural areas east of Siem Reap near the ancient reservoir systems or along rural highways in central Cambodia.
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.