Released: 2026-06-23
Key:
Score Threshold:
google/gemini-3-flash-preview
Score: 4421
Distance: 246km
Lat: 23.791552
Lng: 91.272101
The image shows a red electric rickshaw (toto) common in India and Bangladesh. The overall landscape, utility pole, and architectural style of the white modern building suggest the Agartala region in Tripura, India, or nearby Bangladesh. The presence of ’40’ speed limit signs in that specific font and the general vegetation are characteristic of the outskirts of Agartala.
meta-llama/llama-4-maverick
meta
Score: 4405
Distance: 253km
Lat: 23.7275
Lng: 90.4064
The image depicts a dirt road with tuk-tuks and a speed limit sign displaying ’40’, indicating a developing country with a relatively low speed limit. The presence of tuk-tuks and the style of the buildings suggest that the location is likely in South Asia. The specific architectural style and the surrounding environment point towards Bangladesh, with coordinates near Dhaka.
anthropic/claude-sonnet-4.6
anthropic
Score: 2023
Distance: 1809km
Lat: 13.0827
Lng: 80.2707
The image shows a dirt road with auto-rickshaws (tuk-tuks) parked on the left side, multi-story white concrete buildings under construction on the right, and dry/arid vegetation. The architectural style, vehicle types, and overall environment strongly suggest South India, likely Tamil Nadu or Andhra Pradesh. The auto-rickshaws and building construction style are typical of smaller towns in this region. Given the dry landscape and construction patterns, this appears to be a town in Tamil Nadu or Andhra Pradesh, India. I’ll estimate a location in coastal Andhra Pradesh area.
anthropic/claude-haiku-4.5
anthropic
Score: 2255
Distance: 1593km
Lat: 17.385
Lng: 78.4867
This image shows a typical residential street in India with several distinctive features: unpaved/partially paved road, pink/light colored residential buildings typical of Indian construction, a blue auto-rickshaw on the left, overhead power lines, tropical vegetation including palm trees, and the hazy atmospheric conditions common in certain regions of India. The architectural style, road conditions, and vehicle types strongly suggest this is from a tier-2 or tier-3 city in India, likely in Telangana or Andhra Pradesh region based on the building style and landscape. The image appears to be from a semi-urban or developing area. Based on these characteristics, this appears to be from the Hyderabad metropolitan area or nearby regions in Telangana.
openai/gpt-5-image-mini
openai
Score: 1690
Distance: 2170km
Lat: 10.594
Lng: 103.552
Tropical coastal setting with palms, dusty unpaved street and low concrete-walled guesthouse — typical of smaller coastal towns in Cambodia (Sihanoukville/Otres area). The small open-sided passenger vehicles resemble Cambodian tuk‑tuks/songthaews and the triangular red-bordered warning sign fits regional signage. Overall built form, vegetation and vehicle types point to southern Cambodia near Sihanoukville (around Otres Beach).
moonshotai/kimi-k2.5
moonshot
Score: 718
Distance: 3880km
Lat: 10.4
Lng: 123.9
The image shows several distinctive clues pointing to the Philippines: 1) The red motorcycle-with-sidecar vehicle (tricycle) on the left is iconic public transportation in the Philippines, distinct from the auto-rickshaws found in India or tuk-tuks in Thailand; 2) The tropical vegetation and dusty unpaved road are characteristic of Philippine secondary roads; 3) The modern white concrete building on the right matches recent construction styles in mid-sized Philippine cities; 4) The overall peri-urban setting with utility poles and sparse development suggests a rapidly growing area in the Visayas region (Cebu area) or similar latitude. The coordinates place this in the general Cebu City/Mandaue area of the Philippines, though the exact street cannot be pinpointed without additional landmarks.
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.