Market news feed latency benchmark

How many seconds does a market-moving story take to go from its publish timestamp to actually visible in a public news feed? This page polls 11 free feeds every 60s and measures it. Live, no signup, no email.

Fastest public feed measured so far: SEC EDGAR 8-K at a median 59s from publish timestamp to visible.

63 polls · running 66m · updated 2026-08-13 18:59:12 UTC · raw JSON

Measured publish → visible lag

FeedTypeMedian lagBestWorst Newest item ageSample
SEC EDGAR 8-Kregulatory59s38s80s9munreachable (TimeoutError)
PR Newswirewire2m12s3m4m10 stories
GlobeNewswirewire2m1m2m28m6 stories
Seeking Alphamedia2m56s6m4m48 stories
Business Wirewire8m6m10m55m2 stories
Investing.commedia9m2m32m6m20 stories
Yahoo Financemedia12m7m25m20h9 stories
Nasdaq Trading Haltsregulatory———14hwarming up
Federal Reservemacro———3hwarming up
CNBC Marketsmedia———2hwarming up
MarketWatchmedia———1hwarming up

Lower is better. "Newest item age" is how stale the freshest story in that feed is right now — a quiet desk and a slow desk look the same on that column, which is why the lag columns are the honest ones.

The practical takeaway

Free public feeds are not a speed product. They are a coverage product, and the table above is the price you pay for them: minutes, not seconds, and you have to watch a dozen of them to get the coverage of one desk. If you are trading halts, 8-Ks, guidance cuts or wire-driven gaps, that gap is the entire trade.

If you want the other end of that trade-off — 40+ sources merged into one stream with sub-2-second latency, sentiment tagging and per-ticker watchlists — Top Tier Newswire is the feed built for it, and the delayed tier is free to try. Run it side by side with this table and judge it yourself.

Try the live feed free →

Live merged stream (all 11 public feeds)

Put the live tape on your own site (free)

One iframe, no script, no key, no tracking. It renders the merged headline tape above and updates itself — useful on a trading blog, a Discord landing page or a watchlist dashboard.

<iframe src="https://smeltworks.com/newslag/embed"
  width="100%" height="360" frameborder="0"
  title="Live market headlines"></iframe>

Preview the widget →

Method & FAQ

What exactly is being measured?

For every story we see for the first time, we subtract the publisher's own timestamp on the item from the moment it became visible in the public feed. That difference is the distribution lag a retail trader actually eats. We report the median across all stories seen since this page started polling.

Why is the median lag often minutes, not seconds?

Public RSS is a batch product. Most media feeds are rebuilt on a cache cycle, so a story stamped 14:31:02 can become visible at 14:36. Regulatory feeds (exchange halts, EDGAR) are usually the tightest; consumer media feeds are usually the loosest.

How accurate are these numbers?

Poll resolution is 60 seconds, so each measurement is an upper bound accurate to roughly one minute, and clock skew on the publisher's stamp adds noise. It is deliberately conservative: real lag is at most what is shown, never more.

Does a faster feed actually matter?

It matters where the reaction window is short: halts and resumptions, 8-K drops, guidance cuts, FDA and M&A wires. If your feed is five minutes behind, the first move is over before you read it. For position trading on a multi-day horizon it matters far less.

Built by Smeltworks. Numbers are measured, not vendor-supplied; anything we have not measured yet is shown as blank rather than guessed.