STATIC JSON · NO KEY · NO SERVER

Open data for consumer electronics

Structured specs & computed scores for smartphones, SoCs, GPUs and CPUs — served as plain static JSON. Free and open.

~/techapi LIVE
loading
00 - History

Dataset growth over time

Every dump commit is replayed into a growth chart — from the first snapshot to the latest sync — showing how many records each one added.

v1/index.json
Current snapshot
loading
Growth timeline
Loading growth chart...
  1. Loading repository history...
01 — Playground

Run a request, get JSON

Pick a resource and a slug. The response is fetched straight from the static dataset, right in your browser.

GET /v1/{resource}/{slug}
GET /v1/smartphones/galaxy-s26-ultra
Loading…
04 — How it works

Curated → replayed → served

[ 01 ]

Curated data

Specs live as reviewed JSON files in the repo, each carrying at least one source URL.

[ 02 ]

Replayed to disk

The real API runs in-process and every response is written out — byte-for-byte identical.

[ 03 ]

Served from a CDN

No server, no cold starts. Just fetch() a file — exactly what this page does.

05 — Verification

Checked against reality, not just well-formed

Passing validation only means a record is shaped right. A separate verification layer scores how confident we are that it describes a real, existing device — and only then sets its verified flag.

data/_verify/status.json
records verified
loading…
green yellow red
  1. T0

    Offline trust score

    Completeness, cross-field consistency and source authority score every record into a green / yellow / red band — deterministic, no network.

  2. T1

    Source liveness

    Each record's source_urls are checked for reachability.

  3. T2

    External cross-reference

    Claims are matched against Wikidata by exact title and release year.

  4. T3

    Promotion

    A green record becomes verified; an external contradiction vetoes it. Accuracy stays reality-based.

curl techapi.dev/v1/...

Start fetching in one line

No sign-up, no rate limit, no SDK. Point fetch() or curl at a URL and you have structured specs. Code is MIT, data is CC-BY-SA 4.0.