Kakdwip

City in West Bengal, India

Overview

Kakdwip is a small riverside town in West Bengal, India, known for its proximity to the Sundarbans and the Hooghly River. It offers an authentic taste of rural Bengali life, busy local markets, and is a gateway to several delta islands.

Geographic Context

Jurisdiction
India
State
West Bengal
Population
19,368
Time Zone
Asia/Kolkata
Coordinates
21.88°, 88.18°

Key Metrics

Safety Score · estimate
65
Female Safety
50
LGBTQ+ Safety
40
Traffic Safety
55
Food Safety
55
Monthly Cost · estimate
$340
Local Cost
$205
Local Income
$120
Cashless
45
English
30
Temperature
26.2°C
Summer Avg
29.0°C
Winter Avg
20.3°C
Record High
44.3°C
Record Low
7.4°C
Humidity
75%
Precipitation
1717 mm
Internet Speed
14 Mbps
Upload
4 Mbps
Latency
36 ms
City Vibrancy
—
Coworking
0
Cafes
0
Nightlife
0
Parks
0

Neighborhoods

Estimated

Neighborhood highlights with vibe, safety, and cost profiles.

View featured neighborhoods

Data sources

Source-backed data

World Air Quality Index - Latest available city-level air quality station measurement. Record updated 2025-12-05.
Ookla Open Data - Internet speed aggregates for 2024Q3 from 4 tests. Record updated 2025-12-06.
OpenStreetMap - Nearby amenities counted from OpenStreetMap-derived city extracts. 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 2026-03-22.
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: 48/100; this is not a measured accuracy score. Model: derived+gpt-4.1. Record updated 2026-03-26.