The short answer
The Meta Ad Library shows every active ad but not which one is winning, and sorting by oldest ad points you to the wrong one most of the time. On our board, for 41 of 47 brands with ten or more winning ads, the longest-running ad is not the one with the most estimated spend. Paid alternatives add the missing signal: estimated spend (Ad Radar), saved boards (Foreplay, Swipekit) or a filtered live library (GetHookd).
What the free Library gives you, and what it doesn't
The Meta Ad Library is free, needs no login and covers every active ad on Facebook, Instagram, Messenger and the Audience Network. For each commercial ad you see the creative, copy, start date, platforms, active status and how many versions share the creative.
What you don't get:
| Gap | Why it matters | Who fills it (checked October 2026) |
|---|---|---|
| No spend or impressions for commercial ads | You can't tell a $5K test from a $5M winner | Ad Radar (estimated spend on every ad) |
| No performance sort | You scroll, then guess | Ad Radar, GetHookd, Foreplay (sort by time running) |
| Inactive commercial ads drop out of view (outside EU-delivered ads) | Last quarter's winner is gone when you need it | Foreplay, Swipekit, Atria (saved copies) |
| No tags | You can't filter by hook, angle or format | Ad Radar (ten fixed tags), Foreplay and Atria (AI tags) |
| No transcripts | The spoken hook is invisible to search | Ad Radar, Foreplay, Swipekit, GetHookd |
| No alerts | You find a competitor's new winner weeks late | Ad Radar (winner alerts), Foreplay Spyder, Swipekit Brand Tracker |
Meta does show extra data for one category: for ads about social issues, elections or politics it adds spend, reach and funding entities, and keeps those ads for seven years. That is a transparency feature for political ads, not a research tool for DTC.
The proxy problem: oldest is rarely biggest
Without spend, most people sort by start date and assume the oldest ad is the winner. It is a reasonable guess. Our data says it is usually wrong.
We took the 47 brands with ten or more ads on our board (all live, all with $50k or more in estimated spend) and compared each brand's longest-running ad with its highest estimated spend ad.
6 of 47
brands where the oldest ad is also the top-spend ad
6.7x
median gap between a brand's top ad and its median ad
69
ads under 90 days old already above $1M estimated spend
338
ads live a year or more
In 41 of 47 brands, the ad you would pick by sorting on start date is not the one carrying the budget. Days live is still a useful filter (our post on how long winning ads run explains why), but it filters out losers. It doesn't rank winners. Two pairs from the board show the gap.
I was heartbroken, ready to accept my dog's decline as just "normal aging"
- Est. spend
- $637K
- Days live
- 338
- Format
- Video, 52s
- Variants
- 2
The brand's oldest ad on the board, sending traffic to a quiz. Sort by start date and this is the one you study.
I was at my wit's end and ready to accept my dog's decline as just "normal aging"
- Est. spend
- $15M
- Days live
- 55
- Format
- Video, 52s
- Variants
- 6
Same story, almost the same words, 55 days old, six variants and about 23 times the estimated spend. The landing page changed from a quiz to a listicle. The product claims are the advertiser's.
The difference between the two Wuffes ads is not the creative. It is the destination. That is invisible in the Ad Library unless you click through both, which is why we track the landing page behind each ad as one of the ten tags.
Here's what happens to your body after just four weeks in HF Stride shoes.
- Est. spend
- $252K
- Days live
- 413
- Format
- Video, 86s
Over a year live. The safe pick if all you can see is the start date.
I once followed two women, both 74, with nearly identical health on paper.
- Est. spend
- $889K
- Days live
- 55
- Format
- Video, 363s
A six-minute story ad, 55 days old, with more than three times the estimated spend of the older one. A long-form test that is already scaling.
The alternatives, by the gap they fill
For a spend signal: Ad Radar
Only Meta ads with $50k or more in estimated spend, modeled on engagement tracked over time and labeled as an estimate. Each ad also shows days live and variants, and carries ten tags. $49 every 4 weeks at the launch price (list $99). Meta only, no Chrome extension yet. How any spend estimate works, and where it breaks, is in estimating competitor ad spend.
For saving and sharing: Foreplay or Swipekit
Both fix the "inactive ads vanish" problem by keeping a copy. Foreplay starts at $59 a month ($49 billed annually) with a Chrome extension, mobile app and saves from Meta, TikTok, LinkedIn and YouTube Shorts. Swipekit starts at $26 a month with 5 seats, 5 tracked brands, API and MCP access on every plan, and a 14-day trial with no card. More in swipe file tools.
For a filtered live library: GetHookd
22 million+ active Meta ads with filters by niche, format and runtime, brand tracking and Shopify store data. It deliberately shows no spend figures. $29 a month or $19 billed yearly. See Ad Radar vs GetHookd.
For research plus your own analytics: Atria
Atria follows competitor brands, connects your own ad accounts and generates creative, from $129 a month billed annually. Worth it if you want one tool for both jobs. Our Atria alternatives page lists the trade-offs.
For developers: the Ad Library API
Free, but per Meta's transparency center it covers ads about social issues, elections or politics plus ads delivered to the EU. If your competitors sell only in the US, the API will not give you their commercial ads with performance data. Our Meta Ad Library API guide covers what it returns.
Where the free Library still beats every paid tool
- Coverage. It holds every active ad, not a sample. Paid tools, ours included, track a subset of advertisers.
- Freshness. A new ad shows up in the Library first. Every tool, including Ad Radar, reads from it after the fact.
- Political and issue ads. Real spend ranges, reach and funding entities, kept for seven years. No commercial tool has better data for that category.
- Cost. Zero. For two or three competitors and an hour a week, it is enough.
A free-first workflow
You don't need to pay for anything to start. Pay when the free steps become the bottleneck.
- Search by landing domain, not only by page name. Brands run ads from creator and partner pages. The Ad Library search tips post shows how.
- Filter to active ads, then note start date and version count for each creative in a sheet.
- Click through every landing page. As the Wuffes pair shows, the destination can matter more than the creative.
- Revisit weekly. Inactive commercial ads drop out of view, so save anything you may want later.
- Upgrade when ranking is the problem. If you have 40 candidate ads and no way to tell which one carries the budget, that is the moment a spend signal pays for itself.
Figures marked as estimated spend come from Ad Radar's model of engagement on public Meta Ad Library ads. They are estimates, labeled as such, and are best used to rank ads against each other.