sunglassesguides

Sunglasses Specifications and Comparison API

Structured sunglasses variant, size and lens facts, each with its source and check time, for filtering and comparison in the US market. Built first for AI assistants and agents; people can use the catalog too.

Machine entry points

Market
US only; prices in USD (none covered yet); sizes in millimetres; en-US
Access
Read-only. Anonymous trial without login (30 requests/minute); optional API keys
Data version
c1.o0 (production), published 2026-09-25T08:39:44Z

What is covered right now

81 published variants across 40 models from Ray-Ban (30), Oakley (26), Maui Jim (25). Counts are computed from published records. Target: 80 variants. Offers: not covered (no approved price source connected). Specs are checked manually (target monthly); each record shows its own check time.

FieldKnownUnknownConflicting
lens_width_mm78 (96%)30
bridge_width_mm78 (96%)30
temple_length_mm48 (59%)330
manufacturer_fit_label39 (48%)420
polarized44 (54%)370
vlt_percent19 (23%)620
lens_material29 (36%)520

Tools

Tool (MCP) / operationIdRESTUse it to
get_sunglasses_coverageGET /api/v1/coverageDescribe what this service covers right now: brands, models, published variants, field completeness, offer coverage, freshness and which demo tasks have verified answers.
resolve_sunglassesPOST /api/v1/resolveTurn what a user knows (brand, model name, code on the temple, size marking, lens) into exact catalog variants, or say precisely which detail is missing.
find_sunglassesPOST /api/v1/searchFilter the covered catalog by a reference size with per-dimension tolerances and hard conditions (polarized, maker-labeled bridge fit, brand, optional offer conditions). Every condition is evaluated per candidate as pass/fail/unknown/conflict/stale with evidence.
compare_sunglassesPOST /api/v1/compareSide-by-side comparison of 2-4 variants: every field with its status and sources, which fields differ, conflicts, and size differences against a reference.
get_sunglasses_variantGET /api/v1/variants/{variant_id}Full record of one exact variant: every field with value, field_status, evidence excerpts, sources, check times and unknowns.
get_sunglasses_offersGET /api/v1/variants/{variant_id}/offersReturn covered US offers for one exact variant with item price, shipping, tax, landed-total status, availability and observation time, each kept separate.

Good for

  • Finding frames near an existing size such as 52-18-140 (the numbers printed inside the temple arm), with per-dimension differences
  • Requiring maker-stated polarized lenses or a maker-labeled low bridge fit, with evidence
  • Identifying an exact variant when a model name has several sizes or lenses
  • Comparing 2-4 variants field by field, with unknowns and conflicts shown

Not for

  • Fit guarantees, face measurement, virtual try-on or prescriptions
  • Prices, stock or lowest-price claims (no offer source yet)
  • Purchases or any other action (read-only; Action layer not supported)
  • Brands or models outside the catalog; "no match" is not a market-wide claim

Try it

Find by size

Lens-bridge-temple in mm, as printed inside the temple arm. Type the square symbol as a dash or a space.


Identify a model or code

Examples: "Holbrook OO9102", "Maui Jim Red Sands", "Ray-Ban Original Wayfarer".

Demo tasks and whether they have verified answers

TaskToolCurrent resultReady
Polarized alternatives for an existing 52-18-140 framefind_sunglassesMATCHES_FOUND (3)yes
Identify an ambiguous model nameresolve_sunglassesNEEDS_VARIANT (5)yes
Compare standard and maker-labeled low bridge versionscompare_sunglassesCOMPARED (2)yes
Maker-labeled low bridge fit AND polarized near 52-18-140find_sunglassesMATCHES_FOUND (3)yes
Composite demo plus a 150 USD item-price budgetfind_sunglassesNO_OFFER_COVERAGE (0)no: lens_width_mm unknown for 3/81 published variants

Computed live from the published catalog. The budget task stays not ready until a price source is connected.

Sources and freshness

Facts come from maker pages (Ray-Ban, Oakley, Maui Jim US sites), extracted automatically and recorded only when two independent extractions agree. Nothing is measured by this site. Unknown values stay unknown. See data quality and methodology. Human catalog: browse all variants.