CCTV Camera Database

An open, structured database of 6,325 CCTV / IP camera models and their technical specifications, covering 127 brands across every market segment — from budget consumer WiFi cameras to enterprise PTZ domes and thermal imaging systems. Each camera is a validated JSON file, aggregated into a single queryable dataset (JSON + CSV).

📊 Full field-coverage report → COVERAGE.md

_{cctv-database.com — search 6,325 cameras · full spec sheets · copy-paste Frigate configs · open & auditable}

Camera spec sheets are scattered across vendor PDFs, retailer pages, and paywalled databases (IPVM, etc.) in inconsistent formats. This repo normalises them into one machine-readable structure so they can be compared, filtered, and reused.

The dataset is CC0 and always will be — free to use, copy, and redistribute with no restrictions. The website is just a convenient viewer; the data here is the source of truth.

Every spec is transcribed from a manufacturer datasheet or official product page — never guessed, never inferred. When a field isn't documented it's left empty rather than filled with a plausible value: a missing spec is honest; a fabricated one is a landmine for anyone choosing a camera by it.

• Datasheet-first — retailer/reseller listings only corroborate; they're never the sole source for a spec a datasheet would carry.
• Omit, don't guess — sentinel values (e.g. "0 lux with IR"), marketing rounding, and models with no verifiable spec sheet are dropped, not recorded.
• Traceable — every record carries a sources array of URLs and a last_verified date, so any value can be checked at its origin.

Errors are still possible in any compiled dataset — always confirm against the official datasheet before purchasing or deploying.

The whole dataset is committed and ready to use — no build step, no API key, no rate limit.

const cameras = require('./data/cameras.json');

// All 4K PoE outdoor cameras
const poe4k = cameras.filter(c =>
  c.power_source?.includes('poe') &&
  c.resolution.megapixels >= 8
);

// All cameras with color night vision
const colorNight = cameras.filter(c =>
  c.night_vision?.type === 'color'
);

// All cameras for the UK market
const uk = cameras.filter(c =>
  c.markets?.includes('UK')
);

// All cameras with no subscription fee
const noSub = cameras.filter(c =>
  c.features?.some(f => f.toLowerCase().includes('no subscription'))
);

Or open data/cameras.csv in any spreadsheet for a quick browse.

A read-only, static JSON API is hosted alongside the site (CC0, no key, no rate limit):

For bulk use, grab the whole dataset from data/cameras.json / data/cameras.csv, or a versioned, checksummed cameras.zip from any release.

Browse the database → cctv-database.com

Prefer to self-host or browse offline? A standalone demo (just demo.html + cameras.json, no build step) is included — serve the docs/ folder locally with any static server, e.g. python3 -m http.server inside docs/, then open it.

What you see above:

• Search — instant full-text search across brand, model, and features
• Filter — narrow by brand, camera type, night vision, resolution, or market
• Sort — click any column header to sort ascending/descending
• Detail drawer — click a row to slide open the full spec sheet (resolution, connectivity, protocols, storage, audio, source links)
• Community notes — user-reported quirks and behaviors per camera, sourced and attributed, kept separate from datasheet specs
• Pagination — page through all 6,325 cameras, 25 per page
• Stats bar — live counts for total cameras, brands, 4K+, WiFi, and no-subscription models

Each camera JSON follows schema/camera.schema.json. Required fields:

{
  "id": "reolink-rlc-823a",
  "brand": "Reolink",
  "model": "RLC-823A",
  "type": "bullet",
  "resolution": { "megapixels": 8, "label": "4K UHD" }
}

Common optional fields:

See CONTRIBUTING.md for the full field reference and naming conventions.

Three paths — pick the one that fits you:

The wizard flow:

git clone https://github.com/YOUR_USERNAME/cctv-camera-database.git
cd cctv-camera-database && npm install
npm run add      # asks questions, writes the JSON file
npm run build    # validates everything

See CONTRIBUTING.md for the full field reference, naming conventions, and what makes a good entry.

The data is committed and validated in CI — consumers never build it (see Get the data). These tools are for maintainers and integrators.

npm install   # installs Ajv for schema validation (no runtime deps)
npm run build # validates all JSON, writes data/cameras.json + data/cameras.csv

Other scripts: npm run add (interactive add-a-camera wizard) and npm run check-sources [brand] (probes every sources URL for dead/moved links — see docs/check-sources.md; also runs weekly in CI).

data/rtsp-patterns.json is a CC0 brand-level RTSP URL reference for 172 brands (118 verified / 54 unverified / 349 stream templates). Each path is confirmed against the manufacturer's own documentation — never copied from aggregators. Regenerate after editing strix/verified/:

RTSP_PATTERNS_DATE=$(date +%Y-%m-%d) node scripts/build-rtsp-patterns.js

tools/qa.html is a zero-dependency, offline browser tool for previewing and sanity-checking camera data before you open a PR — a wide grid view across the dataset, gap highlighting, outlier flags, and in-browser schema validation (the schema is compiled to tools/qa-validator.js via node tools/build-qa-validator.js, regenerate after any schema change). Open it from the repo root (it reads ../data/cameras.json, or use the file picker). See tools/plan-demo-and-tools.md.

The shipped configs.frigate snippets are generated; run one of these cameras and you can mark it community-verified (configs.frigate.verified / tested_by / tested_version) via the Verify a Frigate config issue form.

Each configs.frigate block ships a working detect/record baseline, plus structured fields for the deeper integrations: autotracking, onvif_port (the ONVIF control port for manual PTZ — 80 on most cameras, 2020 on Tapo, 8000 on Reolink), and two_way_audio (Frigate live-view talk via the go2rtc backchannel, set where the datasheet confirms two-way audio). The go2rtc methods, per-brand port table, and autotracking rules are documented once in docs/frigate-beyond-baseline.md rather than repeated per camera.

scripts/gen-netbox.js exports cameras into netbox-community/devicetype-library YAML device types — PoE interface (with the standard parsed from power.method), DC power port, microSD module bay, and weight, sourced from each camera's datasheet.

node scripts/gen-netbox.js --brand Hikvision            # → netbox-out/device-types/Hikvision/*.yaml
node scripts/gen-netbox.js --brand Hikvision --report   # list PoE cameras still missing a link speed

A device type is only emitted when the camera has a known RJ45 link speed (network.ethernet_speed_mbps), since the library requires an interface type and it is never guessed. Before opening a PR against devicetype-library, run the generated files through that repo's yamlfmt pre-commit hook so they match its canonical formatting.

Cameras carry a markets[] tag (ISO country codes + region tags like EU/MENA/LATAM) where a regional availability is known. Coverage is intentionally sparse — most entries are untagged or global — so rather than a hand-maintained table that drifts, filter by market on cctv-database.com or query the field directly:

const uk = require('./data/cameras.json').filter(c => c.markets?.includes('UK'));

cctv-camera-database/
├── cameras/              # source of truth — one JSON file per camera, grouped by brand
│   ├── hikvision/        # 479 cameras
│   ├── i-pro/            # 369 cameras
│   ├── dahua/            # 367 cameras
│   ├── acti/             # 251 cameras
│   ├── bosch/            # 231 cameras
│   ├── novus/            # 164 cameras
│   └── …121 more brands
├── data/                 # GENERATED — do not edit by hand
│   ├── cameras.json      # all 6,325 cameras as one array
│   ├── cameras.csv       # flattened, spreadsheet-friendly
│   └── rtsp-patterns.json  # CC0 brand-level RTSP URL layer (172 brands)
├── strix/
│   └── verified/         # per-brand RTSP source files → rtsp-patterns.json
├── schema/
│   └── camera.schema.json
├── scripts/
│   └── build.js          # aggregates + validates cameras/ → data/
├── tools/                # local QA tool (qa.html) + browser schema validator
├── docs/
│   ├── glossary.md       # field reference (source — edit when adding fields)
│   ├── cameras.json      # GENERATED — aggregate copy for the demo/Pages API
│   └── cameras/          # GENERATED — one browsable .md spec sheet per camera
├── CONTRIBUTING.md
└── LICENSE

The dataset and its tooling were assembled with Claude Code. The demo video was produced with hve-spielberg, a Claude Code skill for AI-powered video production built on Remotion.

Shipped: JSON Schema CI, the web frontend (search / filter / sort / compare), Frigate + Home Assistant config export, the NetBox device-type export, and the read-only API above.

Open — contributions very welcome:

• Re-source ~721 reseller-only cameras to official OEM datasheets (#164 / #165)
• Backfill pixel resolution for the ~179 entries that state megapixels but no width×height (#169)
• Frigate verification drive — only 25 of 2,827 shipped configs are community-verified; confirm one you run
• Grow community_notes — the per-camera quirks/behaviors layer (#297)

Tagged releases ship a versioned, checksummed cameras.zip (JSON + CSV) — see the latest release or the full history. The CHANGELOG records what changed in each version. For the always-current data, use data/cameras.json on the default branch.

Specifications are compiled from manufacturer datasheets and reputable retailers and may contain errors or become outdated. Always confirm against the official datasheet (linked in each entry's sources array) before purchasing. Not affiliated with any manufacturer.

If you use this dataset, a link back is appreciated:

CCTV Camera Database (v2.12.0), CC0 1.0. https://cctv-database.com — https://github.com/ch-bas/cctv-camera-database

Data is released under CC0 1.0 (public domain). Trademarks and model names belong to their respective owners.