Castejón de Sos

City in Huesca, Spain

Overview

Castejón de Sos is a charming village set in the Pyrenees of Huesca, Spain, best known for its spectacular mountain scenery and paragliding adventures. Its small population ensures an authentic local experience, with friendly residents and a laid-back atmosphere.

Geographic Context

Jurisdiction
Spain
Province
Huesca
Population
868
Time Zone
Europe/Madrid
Coordinates
42.51°, 0.49°

Key Metrics

Safety Score · estimate
85
Female Safety
80
LGBTQ+ Safety
70
Traffic Safety
90
Food Safety
95
Monthly Cost · estimate
$1,450
Local Cost
$942
Local Income
$1,200
Cashless
50
English
35
Temperature
7.0°C
Summer Avg
15.5°C
Winter Avg
-0.8°C
Record High
30.7°C
Record Low
-19.4°C
Humidity
75%
Precipitation
680 mm
Internet Speed
40 Mbps
Upload
15 Mbps
Latency
59 ms
City Vibrancy
—
Coworking
0
Cafes
0
Nightlife
0
Parks
0
Transport
3
Walk Score
3
Footways
6.0
Airport
—
Distance
—

Data sources

Source-backed data

Ookla Open Data - Internet speed aggregates for 2024Q3 from 1 tests. Record updated 2025-12-05.
OpenStreetMap - Nearby amenities counted from OpenStreetMap-derived city extracts. Record updated 2025-12-08.
OpenStreetMap transport extract - Walkability, footway density, and airport access metrics derived from mapped transport features. Record updated 2025-12-06.

Estimates and unverified origins

Oyster List climate data - Climate origins are unverified. Available methods include historical APIs and AI estimates; stored rows do not identify the original dataset, reference period, or model for each value. Treat these summaries as estimates until independently verified. Record updated 2026-03-23.
Oyster List city price estimates - Price origins are unverified. Available generation methods include AI estimates; stored records do not identify a survey, observation period, or model for each price. Record updated 2026-01-14.
Oyster List AI-assisted estimates - Safety, lifestyle, income, and practical scores include model-assisted estimates. Stored records do not identify supporting observations for each score. Model-reported confidence: 60/100; this is not a measured accuracy score. Model: derived+gpt-4.1. Record updated 2026-03-26.