Elasticsearch alternative

The search engine
that doesn't cost
a fortune to run.

Build a trigram index on any machine — even a $4/month VPS. No JVM. No heap tuning. No cluster. Flatseek indexes your CSV, and you query it via dashboard or CLI. Scale from a single file to billions of rows.

$ curl -fsSL flatseek.io/install.sh | sh

or pip install flatseek · View on GitHub

~60%

lower infrastructure cost vs Elasticsearch

~40%

smaller index size on disk

JVM heap parameters to tune

Cost comparison

Stop paying for hardware
you don't need.

Elasticsearch's JVM heap and cluster requirements add up fast. Flatseek stores everything on disk — memory-mapped, minimal resident set. Your $4 VPS is enough.

Flatseek

Infrastructure $4/month VPS

Long-running node No — query and exit

JVM heap tuning None

Index size Trigram — compact

Dashboard Built-in Flatlens

You only pay for storage.

Elasticsearch

Infrastructure Dedicated cluster, 64GB RAM

Long-running node Required 24/7

JVM heap tuning Required per deployment

Index size Full inverted index, 2–3× larger

Dashboard Separate Kibana license

You pay for cluster + heap + ops.

Two ways to use it

Pick what fits your workflow.

Flatlens for quick file exploration. CLI and Python API for production scale.

Dashboard

Flatlens — for day-to-day CSV work

Open a CSV file in your browser. Search it, chart it, map it. No CLI, no scripts. Built for analysts and anyone who wants answers without a build step.

Drag-drop CSV, JSON, JSONL, XLS
Visual search with filter builder
Bar, line, donut, pie aggregation charts
Interactive Leaflet map with clustering
Encrypt indices with a password

Flatlens docs

CLI & API

CLI & Python API — for production scale

Handle billion-row indices, automate pipelines, embed in your services. Parallel workers, resume on interrupt, REST API for your own frontend.

flatseek build, serve, search, stats, plan
Parallel multi-worker indexing with resume
Cross-index joins on shared fields
Python client — API mode or direct mode
Natural language chat via Ollama

CLI & API docs

Capabilities

Everything you need from a search engine.

Multi-format ingest

CSV, JSON, JSONL, XLS, XLSX. Auto-detect delimiters and column types on import.

Disk-native queries

Memory-mapped trigram index. Resident memory stays low regardless of index size.

Full-text search

Wildcards, fuzzy, phrase match. AND, OR, NOT, grouping. Works on all text fields.

On-disk aggregations

Terms, stats, date histogram, cardinality. Computed without loading docs into RAM.

ChaCha20-Poly1305 encryption

Passphrase-protected indices with PBKDF2-HMAC-SHA256 key derivation.

Parallel builds with resume

Multi-worker indexing with auto-planning. Resume interrupted builds with ETA display.

Try it

From CSV to search in 30 seconds.

flatseek

$ # Build index from CSV

$ flatseek build ./data/people.csv -o ./data

Built 1.2M docs in 12s

$ # Serve API + dashboard — point to your data dir

$ flatseek serve -d ./data

API: http://localhost:8000 · Dashboard: http://localhost:8000/dashboard

$ # Search from CLI

$ flatseek search ./data "city:jakarta AND name:*john*"

DOC {"name": "John Doe", "city": "Jakarta", "email": "john@example.com"}

DOC {"name": "Johnny Silverhand", "city": "Jakarta", "email": "jsilver@example.com"}

$ # Or use Python

$ from flatseek import Flatseek

$ Flatseek("./data").search(q="*john*")

The search engine that doesn't cost a fortune to run.

Stop paying for hardwareyou don't need.