Facebook Ad Copy Length: Limits and How to Test (2026)

The Facebook ad character limit depends on the field and the placement. Meta's Ads Guide recommends 50 to 150 characters of primary text for Facebook Feed and lists 125 for Instagram Feed and Marketplace. Headlines are listed at 27 characters on Facebook Feed and 40 on Instagram Feed and Marketplace. Limits checked 3 November 2026. Confirm each in the Ads Manager preview before you rely on it.

These are Meta's recommended or listed display lengths, not the maximum Ads Manager accepts. The field takes far more text than it displays, and third-party sources quote different maximums, so read the exact cap in the ad editor for your placement instead of relying on any one figure.

Sources: Facebook Feed image, Facebook Feed video, Instagram Feed image and Marketplace image pages of Meta's Ads Guide. The truncation points and the 25-character headline come from Jon Loomer Digital's character-limit chart, last updated 1 March 2025.

Meta's recommendation reaches 150 characters while the independent chart puts the Facebook Feed cut at about 125, so treat 125 as the line to write to.

For Facebook Feed video ads on mobile, Meta says the headline, description and button don't display when the objective is Awareness or the goal is reach, impressions, 2-second views or post engagement. Headline length is moot there. Video seconds are a separate question, covered in our guide to video ad duration limits by placement. For Reels, start with our Facebook Reels ads page.

On mobile, Facebook Feed cuts primary text after about 125 characters and shows a Facebook ad "See more" link, per Jon Loomer Digital's chart, updated 1 March 2025. The opening is the only copy every viewer is certain to see. Everything after it reaches only people who tap, so write the opening as if it were the whole ad. Line breaks can move the cut-off earlier than 125 characters, which is why the ticket below also records characters before the first line break. The checks on this page use the first 125 characters as the unit.

Run this check on every ad. Paste the first 125 characters into a blank note beside the image. Pass: a stranger can tell what you sell and why they should care. Fail: the line only makes sense once the next sentence is read. Our ad hook glossary page lists the hook mechanisms and where each can appear, including primary text.

Our position: settle the first 125 characters before spending a test slot on length. A length test on a weak opening mostly measures the opening, not the length.

Public evidence cannot say. For short vs long Facebook ad copy, there are account-level tests that went both ways, a lifetime dataset with no steady slope, and a snapshot in which brands run both short and long copy. None measures your results. The Facebook ad copy length best practice that holds up is a process: fix the first line, then test length on your own account.

What do public tests show, and what can't they prove?

• AdEspresso, 2 August 2018 (one account's result). Seven ads, about $143 each ($1,000 in total) over 14 days, one ad per ad set (the level where audience and budget are set), aimed at a 1% US lookalike audience (people Meta matches to a customer list) of 2.1 million. The objective was leads for an eBook lead magnet (a free download offered for contact details). Mid-to-long copy had the lowest cost per lead, and the weakest version cost 2.4 times as much per lead as the best. The author's caveat: long-form copy only works if it is written well.
• AdSpyder, 11 June 2026 (observed lifetime, not performance). It sampled 8,815 random Meta ads and measured the gap between the first and last time its tool saw each one. Share still seen 30 days on, by primary text length: 15.5% under 50 characters and 10.8% at 250 to 500, against under 8% across 50 to 250. There is no steady slope. The study says it makes no claim about click-through rate (CTR), cost per click or return on ad spend (ROAS).
• SocialFuel, 31 July 2026 (what brands publish, not what pays). It counted 511 live ads from 17 consumer brands in Meta's public ad library. The median primary text was 153 characters, with the middle half between 106 and 310, and 59.5% of ads ran past 125 characters. Brands run both short and long copy, with brand medians from 28 to 395 characters, and median ad lifespan was flat across length buckets. The post carries no results.

What none of these can prove is which length pays on your account. They differ in objective, audience, year and method.

What advice do you see elsewhere, and why does it conflict?

Four claims recur, and each answers a different question. Hootsuite's post-length cheat sheet, updated 14 July 2026, reports typical Facebook ad lengths from AdEspresso's 2018 analysis of 752,626 ads: a median headline of 5 words and an average primary text of 19 words. Meta's best-practice guidance, as Jon Loomer relays it, is primary text of one to three lines at most. NextAfter's test for the Hoover Institution, run from 15 December 2017 to 22 March 2018 on an ebook offer, found long-form copy converting at 0.50% against 0.35% for short. PPC Hero's 2022 tests for a luxury fashion retailer found that short copy did not win every time, with its edge depending on audience intent. A display rule, typical lengths and two account-level tests answer different questions, which is why they disagree. Treat each as a hypothesis to test, not a target to hit.

Why is a long-running ad not a winning ad?

A long-running ad has only shown that someone left it on. As our guide on long-running Facebook ads puts it, "ad age alone is never proof." Ads stay live as low-spend always-on campaigns, as retargeting or by neglect. Lifetime buckets inherit that weakness. Our inference is that a bucket with higher survival may simply hold more of those ads.

Three things decide which length to test first: how much the offer needs explaining, whether the audience already knows you, and where the ad runs. They point to a test, not to a winner, and we give no character count for any of them.

• Offer complexity. A self-evident offer may need one line. A new, expensive or business-to-business offer may need room for proof.
• Prospecting vs retargeting. Retargeting reaches people who already know you, and prospecting reaches people who don't. See our page on prospecting vs retargeting ads. Our inference is that the two respond to length differently, so run length tests in each separately instead of pooling results.
• Placement. The cut-off differs by placement, so compare lengths only within one placement.

To test Facebook ad copy length, run two ads that differ only in primary text length. Hold everything else equal and judge them on one metric once each ad reaches a minimum number of results that your team sets before the test starts. The test is only as clean as its weakest check.

Name the version before launch. In an after-the-fold test, both ads run past 125 characters and share the same first 125, so the difference reaches mainly people who tap "See more."

In a short-vs-long test, the short ad is under 125 characters and the long ad starts with its exact text, so people who see the long ad read more words without tapping. That second version changes what everyone sees, so read it as a test of the whole message length, not only of the expanded text.

Our guide on how to analyze Facebook ads and choose your next test is the place to start on that minimum. The free Creative Test Capacity Planner shows how many tests your media budget and production capacity support each week. It does not calculate statistical significance.

Copy this ticket into your test log before launch:

• Hypothesis, in one sentence
• Ad A: total characters and characters before the first line break
• Ad B: the same two counts
• Test type: after the fold (both ads past 125 characters) or short vs long (short ad under 125 characters)
• Opening shared: the shorter ad's full text is the opening of the longer ad, and the first 125 characters match if both ads run past 125 characters: yes or no
• Placement and audience
• Metric
• Minimum results per ad
• Decision rule: what ships if A wins, B wins or neither

Length is easy to tag and track in your own log. Record both character counts for every ad, and a later test can search them.

Competitor ads show their copy length but never their results, so treat them as a list of lengths to test, not winners to copy. Build the list in four steps:

1. Group ads by concept first, then count characters for each ad in total and before the first line break.
2. Check whether the first 125 characters stand alone.
3. Record the first-seen date and the observed run time as a hint only.
4. Write down what you cannot see: spend, audience and results.

Rank and run-time data in public ad libraries show how long an ad stayed visible, not whether it paid.

Usually only after offer, hook, format and landing page. Rank length against the other variables by who each one reaches. This ranking is our recommendation, not a measured result.

Whether your queue lives in a spreadsheet or a tool, give length a rank in it instead of treating it as a hunch.

Start by reading your own ads. If the first 125 characters of your primary text cannot stand alone, rewrite them before you touch length. If they can, and offer, hook, format and landing page have already been tested, give length one ranked slot and run it on a single placement with the ticket above.

Choose the test by what you want to learn. If both ads already run past 125 characters, you are testing what people read after "See more." If you are weighing a one-line ad against a long one, you are testing what everyone sees, so judge it on the metric you named before launch. A published average can suggest a length to try, but only your own result can tell you to keep it.