API-First · Schema v1.0
LocRank API
LocRank scores are data first, pages second. The complete index and every benchmark run are published as versioned JSON at stable URLs, so other products can build on the same measurements — with attribution.
Endpoints (v1 — static, keyless, read-only)
| Endpoint | Contents |
|---|---|
GET /directory/api/v1/tools.json | The full tool index: every catalog entry with category, official links, pricing references, affiliate status and benchmark status/scores. |
GET /directory/api/v1/scores.json | The benchmark registry and every run: benchmark versions, corpus versions, dates, per-tool scores and subscores. Append-only — new runs are added, published runs are never altered. |
GET /directory/index.json | The search index consumed by this site's own client-side search (same payload as tools.json in v1). |
Example record
{
"slug": "gemini",
"name": "Gemini",
"category": "ai-translation",
"test": {
"status": "tested",
"score": 99,
"grade": "A",
"benchmark": "Traditional Chinese Quality Benchmark",
"benchmark_version": "1.0",
"run_date": "2026-11-21"
}
}
Versioning
Every payload carries schema_version. Breaking schema changes ship under a new version path (/api/v2/…); v1 URLs keep working. Benchmark scores additionally carry the benchmark name, benchmark version, corpus version and run date — a score without those four fields is not a LocRank score.
Terms of use
- Attribution is required: display “Scores by LocRank” with a link to the tool page or the benchmark run page next to any score you republish.
- Do not present pending tools as scored, and do not strip the conditions disclosure from AI-group scores (see the methodology).
- Reasonable caching is encouraged; the data changes only when a new run or catalog batch is published.
Roadmap (architecture reserved)
- Keyed REST endpoints with per-benchmark and per-tool resources, once demand justifies them — the v1 static schema is the contract they will implement.
- Additional language benchmarks (Simplified Chinese next) publish into the same scores feed.
- A submission endpoint for vendor tool listings (the featured-listing layer) is planned behind human review.
Get API and benchmark updates — and every future score — by email
The public pages show the headline scores. The email edition carries the complete dataset behind them:
- The full per-segment benchmark dataset for every run (all 24 segments, every engine, every flag)
- New-tool test alerts — the day a pending tool gets its first measured score
- Quarterly re-test results, including score changes and what caused them
Free. One list, no spam, unsubscribe anytime. You will receive a confirmation email first.