AI Glasses Specs & Prices API

Clean, current, structured data on every AI and smart-glasses device on the market — without scraping 15 manufacturer sites yourself.

The AI-glasses category is moving fast: new SKUs every quarter, silent price cuts, quiet discontinuations. This Actor maintains a curated, source-verified device database and re-checks prices live against official product pages. Built for AI agents first (MCP-ready, clean JSON) and humans second (CSV/dataset download).

What you get

Every device record carries: brand, model, release date, USD/EUR price, availability, display specs (type, resolution, field of view, brightness), weight, battery, audio, camera, AI features (assistant, live translation, visual AI — the fields generic spec sites don't structure), connectivity, prescription support, companion app, official URL, updated_at, and per-field sources — every value verified against a cited page or explicitly null. No fabricated specs.

Coverage: Ray-Ban Meta (all gens), Meta Ray-Ban Display, Even Realities G1/G2, Xreal One series, Viture, Rokid, Snap Spectacles, Brilliant Labs Frame/Halo, TCL RayNeo, Amazon Echo Frames, Solos, Vuzix — and new Apple/Google entries the moment they ship.

Modes

mode	what it does
list	One row per device: brand, model, price, availability, core flags
specs	Full structured record for one device — fuzzy name match ("ray-ban display" works)
compare	Field-by-field diff of two devices, same flag per row
dump	The entire dataset, every field, with sources
refresh	Live price + availability re-check against official pages, headless browser for bot-walled shops, sanity guards against bogus prices

Example input

{ "mode": "compare", "device_a": "xreal one pro", "device_b": "viture luma pro" }

Example output (compare, excerpt)

{ "field": "display.fov_deg", "xreal-one-pro": 57, "viture-luma-pro": 52, "same": false }
{ "field": "display.brightness_nits", "xreal-one-pro": 700, "viture-luma-pro": 1000, "same": false }

For AI agents (MCP)

Add this Actor as a tool via the Apify MCP server and your agent can answer "which AI glasses under $500 have live translation?" or "compare the Xreal One Pro and Viture Luma Pro" with verified, dated data instead of hallucinated specs. Results land in the OUTPUT key-value record as a single JSON object, and in the dataset for tabular export.

Data honesty rules

• Every field is verified from a fetched source, or null. Never guessed.
• field_sources maps key fields to the exact URL they came from.
• Price refresh rejects implausible scraped prices (bundle pages, accessory hits) instead of silently corrupting records.
• Delisted product pages are flagged (page-gone) — supersession is data too.

Pricing

Pay-per-event: you pay a small fee per device record / compare / refresh run. No subscription.