What Is a Geohash? Encoding Locations for Databases

When you search for "coffee shops near me" on an app and the results appear instantly, there's a good chance a Geohash is doing the heavy lifting behind the scenes. Geohash converts any latitude/longitude pair into a compact alphanumeric string that encodes proximity — and that's exactly what makes it so powerful for databases.

How Geohash Works

Geohash recursively divides the world into a grid of rectangles. At each step, the grid is split in half — first horizontally, then vertically, alternating — and a bit is assigned based on which half the point falls in. After 45 steps (for a 9-character hash), the result is encoded in base32 using the characters 0123456789bcdefghjkmnpqrstuvwxyz.

The result for the Eiffel Tower (48.8584, 2.2945) is u09tunquc. The key insight: two geohashes that share a common prefix are geographically close to each other. All points beginning with u09t are within a few kilometres of Paris.

Geohash in Production: Redis and PostgreSQL

This prefix property makes geohash ideal for database proximity queries:

• Redis has a native GEOADD / GEODIST / GEORADIUS command family that internally uses geohash
• PostgreSQL — you can index a geohash column with a B-tree index and use LIKE 'u09t%' to find all records in a region, orders of magnitude faster than bounding-box lat/lon queries
• Elasticsearch supports geohash grid aggregations natively

Precision Levels

Geohash length controls precision:

• 5 characters ≈ 4.9 km × 4.9 km
• 7 characters ≈ 152 m × 152 m
• 9 characters ≈ 4.8 m × 4.8 m

For most proximity use cases, a 6–7 character hash strikes the right balance between precision and query efficiency.

Limitations

Geohash has one important edge case: two geographically adjacent points can have completely different hashes if they fall on opposite sides of a grid boundary. Production implementations handle this by querying not just the target cell but all 8 neighbours as well.

Use our coordinate converter to generate a Geohash for any coordinate, or our DD to Geohash guide to understand the encoding by hand.