Released: 2026-04-18
Key:
Score Threshold:
meta-llama/llama-4-maverick
meta
Score: 3107
Distance: 951km
Lat: 51.211389
Lng: 12.646111
The image appears to be taken on a rural road in a temperate region, likely in Europe. The road markings, the style of the truck, and the agricultural fields suggest a European country. The language on the road sign is not visible, but the overall scenery is consistent with countries like Germany or Poland. Based on the road layout and the surroundings, a location near Leipzig, Germany is a plausible guess.
google/gemini-3-flash-preview
Score: 4829
Distance: 70km
Lat: 56.402511
Lng: 23.363784
The image displays characteristics typical of the Baltic states, specifically Latvia or Lithuania. The road markings include broken white lines on the edges with diagonal dashes, a style very common in Latvia. The flat agricultural landscape, specific design of the roadside bollards (white with a black band and reflector), and the Volvo truck are consistent with Northern/Eastern European transit corridors. The road geometry and vegetation suggest a location in the southern plains of Latvia.
x-ai/grok-4-fast
x-ai
Score: 2584
Distance: 1320km
Lat: 52.1326
Lng: 5.2913
The image depicts a rural road with flat agricultural fields, dense tree lines, and a straight two-lane road typical of the Netherlands’ countryside. The greenery and cloud patterns suggest a temperate European climate. The truck appears to be a DAF model, common in the Netherlands. Coordinates point to a rural area near Utrecht.
anthropic/claude-haiku-4.5
anthropic
Score: 2595
Distance: 1312km
Lat: 52.35
Lng: 5.25
This image shows a typical Dutch highway scene with distinctive features: flat, wide-open agricultural landscape with green fields extending to the horizon, white dashed lane markings, a truck on the road, vegetation on the roadside, and the characteristic flat terrain of the Netherlands. The road infrastructure, road signs style, and landscape are consistent with a highway in the Dutch countryside, likely in the central or northern Netherlands. The flat terrain and agricultural setting suggest this is somewhere in the Flevoland or North Holland region.
anthropic/claude-sonnet-4.6
anthropic
Score: 3552
Distance: 684km
Lat: 55.6
Lng: 13.5
The flat agricultural landscape, road markings with dashed white lines and yellow center line, the style of road signs, and the Scania/Volvo truck suggest this is in southern Sweden (Skåne region). The flat terrain with grain fields and deciduous trees in the background is characteristic of Skåne. The road style and markings are consistent with Swedish rural roads.
openai/gpt-5-image-mini
openai
Score: 3763
Distance: 568km
Lat: 52.18
Lng: 19.02
Flat agricultural landscape and style of road furniture point to central/eastern Europe. Circular speed/ban signs with red rims and a white sign with a black diagonal (end-of-restriction) are typical EU designs used in Poland. The white roadside delineator posts with black reflective panels, the long dashed centre line and the shoulder rumble markings match Polish rural roads. Vegetation and field patterns also fit central Poland, so I place this on a rural road in central Poland (approx. near 52.18°N, 19.02°E).
moonshotai/kimi-k2.5
moonshot
Score: 3118
Distance: 945km
Lat: 55.8234
Lng: 9.2156
The road markings are the key identifier here. The dashed white lines along the right edge of the pavement (short dashes indicating the carriageway edge) are characteristic of Danish road design. The flat agricultural landscape with expansive green fields is typical of the Jutland peninsula in Denmark. The vehicle is driving on the right side of the road, consistent with Danish traffic rules. The road appears to be a paved two-lane rural highway with European-style signage. The intersection layout with a gravel/dirt side road and the specific style of road signs visible on the right side further support this being Denmark. The vegetation and bright summer lighting also match the Danish climate.
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.