New Research with Claude · Query millions of auction results from your own AI toolsGet Started
Guide · Jewelry & gem appraisers

How jewelry and gem appraisers pull auction comparables and replacement-value evidence

By ALT/FNDATA · October 2026 · 7 min read

Every appraisal needs comparable sales behind the figure, and an insurance replacement value needs evidence a carrier will accept. Gathering comparables by hand across house archives is slow, and it leaves gaps exactly where the stone is unusual. The connector lets you ask for the comparable set in plain language and get back sold lots with the house, the date and a link on each one.

Get your API keyConnect your assistant
YouMarket stats: diamond rings 3ct+AALT/FNDATAIllustrativeMedian $48,000market_stats · 420 soldChristie's · 2024$112,000↗Sotheby's · 2023$63,000↗Bonhams · 2024$29,000↗Every result links to the sale · auction house · date
The aha moment: ask in plain language, get a sourced answer
In short

For colored stones and gem-set pieces, lead with market statistics and whole-market search rather than a single-item price check. Keep loose stones and mounted pieces straight: gems are unmounted, jewelry is finished. Every result links to the sale, which is what makes it replacement-value evidence.

01How it works

Plain-language questions against realized jewelry and gem sales

The connector adds auction price data to the AI assistant you already use. For this work the tables that matter are jewelry (finished, mounted pieces set with stones) and gems (loose, unmounted single stones and parcels). Ask for a segment and the assistant queries it for you, read-only, and returns sold lots with the auction house, the sale date and a source link, normalized to USD.

02The workflow

Building a comparable set for a report

Start with a statistical picture of the segment, then pull the specific comparable rows you want to cite, and reach across the whole market when the piece is unusual.

1

Summarize the segment with market statistics

Ask your assistant“Market stats for emerald-cut diamond rings over 3 carats sold since 2022.”
Toolsmarket_stats
What comes back

Count of sold lots, median, 25th and 75th percentiles, average, minimum and maximum, the sale-date span, and the leading houses by volume for that segment. This is the backbone of a replacement-value argument: the central tendency and the spread, from real sales.

Tip. For colored stones, market_stats over the segment is the right tool. Single-item colored-stone comparables are held back, so a statistical picture is both available and more defensible than one match.
2

Pull the exact comparable rows to cite

Ask your assistant“Show sapphire rings sold above $20,000 since 2023, newest first, with the house and sale date.”
Toolsquery_tableget_table_schema
What comes back

A filtered, sorted list of sold lots with the fields you asked for. You can filter by price, date, gem type, carat and more, sort by any column, and choose the fields to return. Each row carries its auction house, sale date and link, ready to drop into a schedule of comparables.

Tip. Ask “what columns does the jewelry table have?” first if you want to filter on a specific field. That calls get_table_schema and shows you exactly what you can query.
3

Reach an unusual stone across the whole market

Ask your assistant“Search the whole market for a Burma ruby ring over $50,000.”
Toolssearch
What comes back

A keyword search across every category at once, merged by realized price. For a rare origin or an exceptional stone, this surfaces results wherever they sold, including signed pieces catalogued under a different heading.

Tip. If a search of jewelry comes back thin, ask the assistant to try the gems table, and the reverse. A loose stone and the same stone mounted live in different tables.
4

Keep loose stones and mounted pieces straight

Ask your assistant“What loose unmounted sapphires over 5 carats have sold at auction?”
Toolsquery_tablemarket_stats
What comes back

Results from the gems table, which holds loose, unmounted stones, the right source for a gemstone valuation. The jewelry table holds the finished, mounted pieces. If one returns nothing, the other usually has it.

Tip. State “loose” or “mounted” in your question. It steers the assistant to the correct table and saves a round trip.
03Reading the answer

Confidence, caveats and why the number is defensible

For jewelry and gems your evidence rests on the statistical picture and the cited rows rather than a single automated estimate, because colored-stone single-item comparables are held back by design. That is the honest and the stronger position for a report: a segment median with its quartiles, plus a schedule of named comparable sales, each linking to the auction house and date. When you do run a price check on a mounted piece, read its confidence label the same way as elsewhere: a tight, well-specified match reads high, while a broad or scattered pool is flagged and reads low. A carrier can follow every figure back to the sale it came from.
04Honest limits

What the data does and does not cover

Know the boundaries before you lean on the connector for an appraisal.

  • It reports realized auction results normalized to USD, not retail replacement cost or private-treaty prices. Replacement-value work still applies your judgment on top of the market evidence.
  • Single-item colored-stone comparables are held back, so for gemstones and gem-set pieces use market_stats and cited query_table rows rather than one price check.
  • gems is loose stones, jewelry is mounted pieces. Query the right one, and try the other if a result is empty.
  • Grading and certification detail depend on how each house catalogued the lot, so read the returned fields before treating two lots as like-for-like.
  • It is market evidence, not a formal appraisal or personalized financial advice.
05Getting started

Connect it once, then ask in plain language

Access requires a paid ALT/FNDATA API key. Create one in the app, connect it to your assistant using the setup page for your tool, then ask for the comparable sets above.

1Get an API keyPaid access in the ALT/FNDATA app. There is no free tier.2Add the connectorFollow the setup page for Claude, ChatGPT, Gemini, Perplexity or your tool.
3Ask awayUse the prompts above. Every result carries a source link you can cite.

Questions

Frequently asked questions

What is the difference between the jewelry and gems tables?
The gems table holds loose, unmounted stones and parcels, which is the right source for a gemstone valuation. The jewelry table holds finished, mounted pieces set with stones. If a search of one is empty, try the other.
Why can't I price-check a single ruby or sapphire?
Colored-stone single-item comparables are held back. For a stone or a gem-set piece, use market statistics over the segment plus a schedule of cited comparable sales, which is a more defensible basis for a report anyway.
Can I use this for insurance replacement value?
Yes, as market evidence. Each comparable carries the auction house, sale date and a source link, so a carrier can trace every figure. Replacement-value work still applies your professional judgment to the market picture.
Are the prices realized or estimates?
Realized, normalized to USD. Pre-sale estimates and unsold lots are handled separately, so a summary reflects pieces that actually sold.
Is there a free tier?
No. A paid ALT/FNDATA API key is required, created in the app.
Keep reading

Related guides

Valuing client collectiblesWatch comps and price trendsCollector market researchJewelry & gems dataset

Put a sourced number on it today.

Connect the ALT/FNDATA connector to the AI assistant you already use, and pull realized auction comparables, market statistics and price trends in plain language, each with a link to the sale.

Get your API keySee the setup pages

The ALT/FNDATA connector is read-only and returns realized auction results across 20 categories from 650+ auction houses, normalized to USD, each with the auction house, sale date and a source link. It reports what items sold for at auction, not retail or private-treaty prices. Methodology · All datasets · All guides