Overview
The FRED Series listing delivers 574,480 series from FRED (Federal Reserve Economic Data), with 69.9 million observations and their full history, ready to query in Snowflake with no API keys or rate limits to manage. It also includes 826 million ALFRED vintage rows, so you can see every series as it was known on any past date. Start with a free 30-day trial with full access. It’s built for economists, quant researchers, analysts and data teams who need macro data next to their own data in Snowflake: rates, inflation, jobs, GDP, housing, banking and regional data, without writing and maintaining a FRED pipeline. Data is standardized into analytic‑ready views (series, observations, vintages, categories, tags, releases) with metadata on every series (units, frequency, seasonal adjustment, popularity) and automated pipeline quality checks (load parity, unique keys, matching series between views) so you can query, join, model and build dashboards right away.Start your free 30-day trial on the Snowflake Marketplace
- Free 30-day trial with full access to every view, series and vintage
- 574,480 FRED series and 69.9 million observations, from 61 sources such as the Census Bureau, BLS, BEA and the Federal Reserve
- Full history of every series, back to 1854 for the NBER recession indicators and 1948 for the unemployment rate
- ALFRED vintages: 826 million rows covering every published version of each observation, for point-in-time analysis
- Rich metadata: 240 releases, 4,599 categories and 5,567 tags to find and group series
- Refreshed daily: series FRED updated the day before are reloaded each night
- Automated pipeline quality checks (load parity, unique keys, observations matching their vintages)
What is FRED data?
FRED (Federal Reserve Economic Data) is the economic database of the Federal Reserve Bank of St. Louis. It collects time series from government agencies, central banks, international organizations and universities, and publishes them under one ID scheme (UNRATE, CPIAUCSL, GDP). Its companion archive, ALFRED, keeps every past version of each series, so you can see what was known before later revisions. FRED is the standard free reference for U.S. and international macro data.
What’s available, at a glance:
- Labor: unemployment, payrolls, wages, job openings and claims, nationally and by state and metro area
- Prices and inflation: CPI and its components, PCE, PPI and deflators
- Output and income: GDP and its parts, industrial production, capacity utilization, personal income
- Interest rates: Fed funds, Treasury yields and spreads such as
T10Y2Y, mortgage rates - Money and banking: money supply, reserves, bank loans, deposits and balance sheets
- Housing: starts, permits, sales and FHFA house prices
- Regional: state, county and metro data from the Census Bureau, BLS and BEA
- International: OECD, IMF, World Bank and Eurostat series by country
Coverage
As of November 2026, by series frequency:USREC. Some projection series run to 2036. The ALFRED vintage views cover 574,361 of the series, with 23.6 million vintage dates.
What isn’t included: FRED’s terms say series owned by third parties need the owner’s permission to use beyond personal use, so the listing leaves out every series from licensed sources: 96,159 series, including S&P Dow Jones Indices (such as SP500 and DJIA), ICE BofA bond indexes, Moody’s, Nasdaq and Zillow. Every other FRED series is included. See What FRED Economic Data Is on the Snowflake Marketplace? for the full breakdown and alternatives.
Schema
The listing has eight views in theFRED schema. Every view has series_id, the FRED series ID (e.g. UNRATE), so join on it. The listing also includes a semantic view, FRED_ECONOMIC_DATA_SV, and a Cortex agent, FRED_AGENT, for asking questions in plain English.
fred_series_vw
One row per series.
fred_series_observations_vw
The current value of every observation: one row per (series_id, date).
fred_series_observations_vintages_vw
Every published version of every observation, from ALFRED: one row per (series_id, date, realtime_start).
fred_series_vintage_dates_vw
One row per (series_id, vintage_date): each date on which the series was revised or new data was released.
fred_series_categories_vw
One row per (series_id, category_id). A series can sit in more than one category.
fred_series_tags_vw
One row per (series_id, tag_name).
fred_series_releases_vw
One row per series: the FRED release it’s published in.
fred_series_search_vw
One row per series, with one text column for searching by keyword.
Example queries
All example output below is real, pulled from the listing on November 20, 2026 at 03:23 UTC. Each query finished in under 3.5 seconds on an X-Small warehouse. Run the same queries yourself during the free 30-day trial. The queries name only thefred schema, so they work in whatever database you mount the listing as.
What is the latest unemployment rate?
What is the 10-year minus 2-year Treasury spread?
Find series by keyword
The most popular series that mention mortgage rates:Release, category and tags for a series
How was a jobs number revised?
Every published value of June 2026 nonfarm payrolls (PAYEMS, thousands of persons), from ALFRED:
Vintage observations (ALFRED)
FRED revises values after it first publishes them.fred_series_observations_vw holds the current value of each observation. fred_series_observations_vintages_vw holds every published version, from ALFRED, FRED’s archive. Each row is one version of a data point, valid from realtime_start to realtime_end, so you can rebuild a series as it was known on any past date. Use it for backtests and forecast evaluation without look-ahead bias.
GDP (billions of dollars, seasonally adjusted annual rate) as it was known on March 1, 2020:
fred_series_vintage_dates_vw lists every vintage date for each series:
Data Ingestion
Data is ingested into Snowflake with best practices in mind.Find what changed
fred_series_vw.Fetch full history
Load and replace
Vintages, categories, tags and releases
Data Quality
These checks run with every load, and a failure is flagged on that run:- Load parity: every staging load checks that the rows Snowflake loaded equal the rows parsed from the files, with no load errors.
- Every changed series staged: every series selected for reload must end up staged in S3, and the count selected must match the count FRED reported as updated.
- Unique keys: each
(series_id, date)has exactly one current value, and each series has one metadata row. - Matching series:
fred_series_vwandfred_series_observations_vwmust hold the same set of series, so no series has metadata without values or values without metadata. - Vintages agree with current values: for every series with a complete vintage load, the current ALFRED versions must equal the current observations.
- No overlapping vintages: no two versions of the same data point can have overlapping real-time periods, and each data point has exactly one current version.
- Vintage completeness: the staged vintage rows must equal what FRED returned for each series, and the vintage dates must match the ones FRED lists.
- No series is ever emptied: a reload that returns no values for a series leaves its existing values in place.
- When FRED has no value for a date (it publishes
.), there is no row infred_series_observations_vw. In the vintage view, those versions are kept with a nullvalue. - As of November 2026, 114 series have metadata in
fred_series_vwbut no observations yet, and 119 have no ALFRED vintages. - Categories, tags and releases are loaded when a series first arrives. They aren’t reloaded for existing series each day.
fred_series_observations_vw on the next nightly load. The old value stays in fred_series_observations_vintages_vw, with its realtime_end set to the day before the revision.
Disclosures
Fruit Stand Data LLC is not endorsed by or affiliated with the Federal Reserve Bank of St. Louis, FRED, ALFRED or any of the sources whose data FRED publishes. The listing includes only series FRED does not mark as owned by a third party; most are public domain, and FRED asks that you cite the source (see each series’cc tag and notes). Use contact@fruitstand.dev for questions.
F.A.Q.
Frequently asked questions from our clients.Is there a free trial?
Is there a free trial?
What is FRED?
What is FRED?
How often is the data refreshed?
How often is the data refreshed?
When does replication across cloud regions happen?
When does replication across cloud regions happen?
How far back does the historical data go?
How far back does the historical data go?
PAYEMS), 1947 for GDP and CPI, 1948 for the unemployment rate (UNRATE) and 1976 for the 10-year minus 2-year spread (T10Y2Y). The vintage history goes back as far as ALFRED has it for each series.How is the data obtained?
How is the data obtained?
Why use Fruit Stand's FRED Snowflake listing instead of using the FRED API?
Why use Fruit Stand's FRED Snowflake listing instead of using the FRED API?
What does it cost to build this ourselves vs. buy the listing?
What does it cost to build this ourselves vs. buy the listing?
Do you include revision history (vintages)?
Do you include revision history (vintages)?
fred_series_observations_vintages_vw has every published version of each observation from ALFRED, with realtime_start and realtime_end dates, so you can see any series as it was known on a past date. See Vintage observations above.What isn't included?
What isn't included?
SP500, DJIA), ICE BofA bond indexes, Moody’s, Nasdaq and Zillow, because FRED’s terms require the owner’s permission to use them beyond personal use. Some have public alternatives in the listing, such as BAA10Y for corporate spreads and the FHFA house price index. See What FRED Economic Data Is on the Snowflake Marketplace? for details.Can I request additional data from FRED to be made available in Snowflake?
Can I request additional data from FRED to be made available in Snowflake?