How to Download Video From Meta Ad Library (2026, 4 Methods)

By the AdMapix Research Team · Updated April 16, 2026

Downloading for private creative research is generally protected fair use under US copyright law. Redistributing or using a competitor's video in your own ads is not. This guide covers research use only.

Meta's Ad Library doesn't have a native download button. Four methods work in April 2026: (1) Chrome DevTools extract the .mp4 URL, (2) trusted browser extensions, (3) screen recording, (4) the official Ad Library Report CSV for political ads only. Here's each method by risk level — and the legal lane that keeps you on the right side of copyright.

If you landed here because you couldn't find the Library at all, start with our prerequisite guide on where to find the Meta Ad Library. Once you're looking at an advertiser's active creatives and you want the actual .mp4 file on your drive, the rest of this article is for you.

Why there's no native "download" button

Meta launched the Ad Library as a transparency tool, not an asset repository. The mission is to let journalists, regulators, and the public inspect which ads a Page is running — particularly political and issue ads, which is why that subset ships with a downloadable CSV while commercial creatives do not.

Three practical reasons Meta keeps the download button off:

1. Copyright liability. Every video is a brand asset owned by the advertiser. A download button would put Meta in the middle of endless takedown disputes.
2. Scraping deterrence. One-click export makes mass harvesting trivial. Meta's Terms of Service prohibit automated scraping at scale; the missing affordance is part of that posture.
3. Competitive abuse. Meta knows people will reupload competitor creatives into their own ad accounts. Friction discourages casual theft.

Good news for researchers: Meta never claimed the videos were technically inaccessible. The <video> element is standard HTML5. The stream lives on video.xx.fbcdn.net. Any browser that can play the ad can fetch the bytes. That's the seam we work in.

The four methods, ranked by risk

We've tested every method that was worth testing. Here's the decision matrix:

How to pick:

• Grab a single ad for a swipe-file? Method 1. It's free, safe, and gives you the original quality file.
• Building a Notion board of 40 creatives an hour? Method 2, but only after reading the extension warning below. Speed comes with risk.
• Paranoid about ToS and only need the visual for a report? Method 3. Legally bulletproof because you're recording your own screen.
• Studying political or issue ads specifically? Method 4. Slower but Meta explicitly sanctions it.

Let's walk through each.

Method 1 — Chrome DevTools (verified April 2026)

This is the method we use ourselves. It takes about 30 seconds per ad once you've done it twice, costs nothing, and leaves zero footprint on your account. Meta shuffled the Ad Library filter UI in January 2026 (country and date-range dropdowns moved to the top-right), but the underlying <video> element and video.xx.fbcdn.net CDN pattern are unchanged — so the walkthrough below still works as of our last test on April 14, 2026.

Step-by-step:

1. Open Meta Ad Library. Go to facebook.com/ads/library, pick your country and ad category (usually All ads), then search the advertiser's Page name.
2. Find the ad you want. Scroll until you see the video creative. Don't click play yet.
3. Open DevTools. Right-click anywhere on the ad card and choose Inspect, or press F12 (Windows) / Cmd+Option+I (Mac). A panel opens on the right or bottom.
4. Switch to the Network tab. In the DevTools top menu, click Network. You'll see a blank or cluttered request log.
5. Filter for media. In the Network tab's filter row, click Media. (If you don't see it, click the filter icon and check the Media box, or type mp4 into the text filter.) This hides everything except video and audio requests.
6. Play the video. Click the play button on the ad. DevTools will log the video request as it loads.
7. Isolate the .mp4 request. You'll see one or more rows — the main one will typically have a URL starting with https://video.xx.fbcdn.net/... and a file extension of .mp4. Click that row.
8. Copy the Request URL. In the right panel, under Headers, find Request URL. Right-click the URL and choose Copy link address, or select it and Cmd/Ctrl+C.
9. Open the URL in a new tab. Paste into a new browser tab and press Enter. The video will play fullscreen on a black background.
10. Save the file. Right-click the video and choose Save video as.... Pick a folder, rename the file (we recommend {brand}_{libraryID}_{date}.mp4 — more on that below), and save.

Fallback if the Network method misbehaves: In DevTools, switch to the Elements tab. Press Cmd+F (or Ctrl+F) inside the Elements panel and search for <video. You'll land on the video node. Look at its src attribute (it'll be another video.xx.fbcdn.net URL). Copy that src value, paste it into a new tab, then Save video as... exactly like step 10. The Elements fallback also gives you the poster attribute if you want the thumbnail image.

Tips we learned the hard way:

• Some ads stream in HLS chunks instead of a single .mp4. If you see .ts segments or a .m3u8 manifest instead of an .mp4, scroll the Network log — there's usually a consolidated .mp4 request fired on first play. Refresh the page and try again before giving up.
• Carousel video ads have a separate <video> node per card. Scroll each card in the carousel, then re-check the Elements tab. Each card's video has its own src.
• Save the ad copy to a .txt alongside the video. In the DOM, the creative text lives in the ad card; select it, copy, paste. Researchers who skip this step regret it six weeks later.

Method 2 — Browser extensions (vet carefully)

Browser extensions compress Method 1 into a single click. The trade-off: you're handing an unknown third party access to every tab you open. In February 2026, CL Suite — an extension with 40k+ Chrome Web Store installs marketed as an ad-library downloader — was caught exfiltrating Meta Business Suite TOTP codes and session cookies. Google pulled it within 48 hours, but dozens of agency accounts were already compromised.

That incident isn't an outlier. Ad-downloader extensions are a popular malware target because they need broad site permissions and their users are logged into high-value ad accounts. The vetting checklist we use:

• 10,000+ installs minimum. Low-install extensions are either abandoned or fresh malware vehicles.
• Verified publisher badge on the Chrome Web Store listing. Anonymous publishers = automatic pass.
• Updated within the last 90 days. Abandoned extensions get sold — the Chrome Web Store has documented trusted extensions bought by bad actors who ship a malicious update weeks later.
• Minimal permission scope. If a downloader wants read and change all your data on all websites plus clipboard and credential access, decline. A legit downloader needs facebook.com/ads/library and fbcdn.net, nothing else.
• Independent review. Search the extension name on security-focused subreddits and Bleeping Computer.

Operational rule: run any ad-research extension in a dedicated Chrome profile that is never logged into Meta Business Suite, your bank, or your email. If the extension turns malicious, the blast radius stays inside a sandbox.

We deliberately don't name specific extensions here. Names that are safe in April may be malicious in June after an ownership transfer. Use the checklist, install only what you'll use this week, and uninstall when you're done.

Method 3 — Screen recording (zero-risk fallback)

If you need legal ironclad-ness, screen recording is the move. You're capturing your own screen displaying publicly available content — same as a researcher with a smartphone camera — and no TOS argument touches it. The cost is quality: you lose encoding fidelity, and the capture is only as clean as your monitor resolution.

Built-in options:

• macOS: Cmd+Shift+5 opens the capture toolbar. Pick Record Selected Portion, draw a box around the ad, hit Record. Files land in ~/Desktop as .mov.
• Windows 11: Win+G opens Xbox Game Bar, click the Record button, capture the active window. Files save to Videos/Captures. Snipping Tool also ships screen-record in 22H2+.
• Cross-platform: QuickTime (macOS) and OBS Studio (Mac/Windows/Linux) give you frame-rate and audio-source control. OBS is overkill for one ad but perfect for 50 in a session.

Quality tips:

• Set display resolution to at least 1080p. A Retina Mac at default scaling records at roughly 540p effective.
• Play the ad in full-card view, not the thumbnail.
• For voiceover, record system audio: on macOS, the built-in recorder only captures mic audio; use QuickTime with BlackHole, or OBS.

Screen recording is our go-to when a client asks "is this legal?" and we don't want the conversation.

Method 4 — Official Ad Library Report CSV + political-ads API

For research that needs an official paper trail, Meta offers sanctioned data access — but only for political and issue ads. The Ad Library Report (CSV, political/issue ads in supported countries) and the Ad Library API (programmatic, developer approval required) are both documented.

What you get:

• Report CSV: spending ranges, impression ranges, demographic reach, time ranges. Seven-year retention for political/issue ads. Great for regulatory analysis, academic research, or FOIA-style journalism.
• API: returns ad_snapshot_url, ad copy, targeting information where disclosed, and limited media metadata. You can programmatically query an advertiser's political ad history.

What you don't get: a commercial-creative feed. Meta's API excludes non-political ads from programmatic media access by design. If your use case is "pull every Allbirds video ad from the last 90 days in one script," the official API will return nothing useful. That's by policy, not by bug.

API approval needs a Meta Developer account, platform agreement, and verified identity. Plan 2-10 business days. For one-off research, the CSV is faster and gives you everything the API does.

Is it legal? Fair use explained

The short version: yes, for private research, usually.

Under US copyright law, fair use lets you reproduce a copyrighted work without permission if four factors (together, not individually) favor you:

1. Purpose and character of use. Non-commercial research, commentary, criticism, and education sit squarely inside fair use. Downloading a competitor's video to dissect the hook structure in a private Notion doc is textbook transformative-ish research.
2. Nature of the work. Ads are commercial creative works, which cuts slightly against fair use — creative works are more protected than factual compilations. But ads are also published with the intent of being seen widely, which softens that factor.
3. Amount and substantiality. You're copying the whole ad. This factor weighs against you, and there's no way around it.
4. Market effect. Does your copy harm the market for the original? If it's in your private folder and you're not redistributing it, the answer is no. This is the factor courts weigh most heavily.

The case law that matters here is Lenz v. Universal Music Corp., 801 F.3d 1126 (9th Cir. 2015) — often called the "dancing baby case." The Ninth Circuit held that fair use is an affirmative right, not merely a defense to infringement, and that copyright holders must consider fair use in good faith before issuing a DMCA takedown. That ruling strengthens the legal posture of researchers who download content for analysis, because it frames fair use as a use you're entitled to, not a plea you fall back on when you get sued.

Two more citations worth bookmarking: the US Copyright Office's plain-English four-factor guide to fair use, and Meta's own copyright policy page for context on how the platform handles infringement claims.

None of this is legal advice. If your use case is anywhere near the redistribution line, talk to an IP lawyer. The guidance in this section is for creative teams doing normal competitive research, not for media companies planning to republish.

What you can't do

The list of things that flip you from fair use into infringement is short and unambiguous:

• Reuploading a competitor's ad to your own ad account — not even "for testing." This is the single most common violation we see agencies commit.
• Using a clip of a competitor's video in your own paid ads. That's derivative commercial use and it's not defensible.
• Publishing a downloaded ad in a public swipe-file, Twitter thread, or newsletter. Private = fair use. Public = distribution = infringement risk.
• Training a generative AI model on downloaded creatives without a license. This is an open legal question in 2026, but the conservative answer is "don't."
• Mass-scraping the Library with automated scripts. Individual browser-based downloads are one thing; running a headless Chrome farm that pulls 10,000 ads overnight violates Meta's Terms of Service regardless of the fair-use status of any single download.

Stay inside private research and you're fine. Cross any of the lines above and the fair-use shield drops.

Build a competitor video swipe-file workflow

Downloading is the first step; organizing is what makes downloads useful six months later.

Notion or Airtable schema we use:

• Brand (select)
• Library ID (text — Meta's unique ad ID, visible under the ad card)
• First seen (date)
• Format (select: video, image, carousel, dpa)
• Hook 0-3s (text — what happens in the first three seconds)
• CTA (select: shop, learn more, sign up, download, other)
• Duration (number, seconds)
• Vertical / niche (select)
• File link (URL — point to cloud storage, not a local path)
• Notes (long text — why you saved it, what hypothesis it tests)

File naming convention: {brand}_{libraryID}_{YYYY-MM-DD}.mp4. This keeps files self-documenting even if your database blows up.

Storage: We use a private Cloudflare R2 bucket with signed URLs. Google Drive and Dropbox also work. Avoid local-only storage — you'll lose the archive the next time your laptop dies.

Review cadence: Once a week, 30 minutes. Tag three creatives with steal-worthy hook, one with test this angle, one with don't-ever-do-this. That's your weekly creative sync material.

For the full cross-platform version of this workflow — covering TikTok Top Ads, LinkedIn Ads Library, and Google Transparency Center — see our cross-platform competitor research SOP. If you're specifically wrangling multi-platform intake, our spy-on-ads-all-platforms walkthrough covers the intake side end to end.

When to skip manual methods and use a paid tool

The honest truth: if you're downloading more than 20 ads a week, the DevTools method starts to feel like tax prep. That's when a paid ad intelligence tool pays for itself.

We built AdMapix's creative vault to skip this download dance entirely. Every ad we surface ships with a permanent R2-hosted media URL, auto-tagged hook/CTA/vertical metadata, and an export path that drops straight into Notion, Airtable, or Feishu without a VPN or a screen recorder. We made that call because our own team was losing four hours a week to the download-rename-upload loop.

We're not the only option and we're not always the right one. If you want to compare, our honest roundup is the best ad spy tools in 2026 — we include competitors who beat us on specific use cases. For most teams the calculus is: below ~20 downloads a week, Method 1 is fine; above that, a paid tool saves more time than it costs.

FAQ

Can I download Meta Ad Library videos legally? For private competitive research, yes — fair use under US copyright law generally covers it. For redistribution, reupload into your own ad account, or use in your own paid ads, no. The line is private analysis versus any form of republication or commercial reuse.

Is the DevTools method still working in 2026? Yes, verified April 14, 2026. Meta moved the filter UI in January 2026 but the underlying <video> element and video.xx.fbcdn.net CDN pattern are unchanged. Chrome DevTools' Network tab Media filter still exposes the .mp4 request as described. Full walkthrough in the Chrome DevTools docs.

Will Meta ban my account for downloading? Individual browser-based downloads from your logged-in account have never been a known trigger for account action. Automated scraping at volume is a different story and explicitly violates Meta's Terms of Service. Keep it to manual, browser-speed downloads and you're fine.

Can I use downloaded competitor videos in my own ads? No. That's derivative commercial use of someone else's copyrighted creative and it is not protected by fair use. Use the download for analysis, then produce your own original video inspired by what you learned.

How do I save ads in bulk? Meta doesn't offer a first-party bulk download for commercial ads — only the Ad Library Report CSV for political/issue ads. For bulk commercial capture, the options are a vetted browser extension (see our warnings in Method 2) or a paid ad intelligence tool with native export. Do not run automated scraping scripts against the Library.

Why won't some videos download? Three common reasons: (1) the ad streams in HLS chunks (.ts + .m3u8) instead of a single .mp4 — refresh the page and re-inspect, the consolidated .mp4 usually fires on first play; (2) the ad is a carousel and each card has its own <video> node you need to isolate; (3) Chrome is caching an earlier request — clear the Network log, hard-refresh with Cmd/Ctrl+Shift+R, and try again.References: Meta Ad Library · Meta copyright policy · Chrome DevTools Network docs · Lenz v. Universal (EFF case page) · US Copyright Office fair-use guide · Chrome Web Store developer policy