How to Search Your Old Liked Tweets (Complete Guide 2026)

Short answer: X has no search for your likes. The search bar searches all of X, not the posts you liked, and the Likes tab only scrolls. To find an old like, scroll the Likes tab (slow, and it stops loading on large histories), download your X archive and search its like.js file, which lists every post you liked back to your first, or load that archive into a tool that searches by meaning.

The archive is the only complete record. It holds each liked post's ID, text and link, but not who wrote it or when you liked it, so every method below inherits those two gaps.

Why X's Search Bar Doesn't Find Your Likes

X's search bar searches public posts across the whole platform. It has no filter for "posts I liked", so a search returns whatever matches the keywords and ranks well, mixed with posts you never saw.

What that means in practice:

• There is no way to limit a search to your own likes
• Matching is by exact keywords, not meaning
• Results lean toward recent, high-engagement posts

The Likes tab itself has no search box at all. It is a single reverse-chronological list.

Method 1: Manual Scrolling Through Your Likes Tab

Difficulty: Easy | Time Investment: High | Reliability: Low

The most straightforward approach is returning to your profile and manually scrolling through your Likes tab.

How to do it:

1. Go to your X profile
2. Click the "Likes" tab
3. Scroll backward through time until you find the tweet

The reality:

If you liked that tweet 3 months ago and you like 20 tweets per day, you're looking at scrolling through approximately 1,800 tweets. On mobile, each scroll loads about 10-15 tweets, meaning you'll need to scroll at least 120+ times.

Pros:

• No additional tools needed
• You can browse adjacent tweets from the same timeframe
• Sometimes triggers memory by seeing related content

Cons:

• Extremely time-consuming for prolific likers
• Easy to lose your place if you refresh accidentally
• X's infinite scroll can be buggy, sometimes jumping or freezing
• No filtering or search capability

Best for: Tweets liked within the last week or when you remember approximately when you liked it.

Method 2: Using X's Search Operators (Limited)

Difficulty: Medium | Time Investment: Medium | Reliability: Medium

X supports advanced search operators that can narrow down results, though they won't directly search your likes.

Useful search operators:

from:username keyword
"exact phrase"
keyword since:2025-01-01 until:2025-12-31
keyword filter:links
keyword min_faves:100

How this helps with likes:

If you remember who posted the tweet or specific keywords, you can combine operators:

from:naval life advice

This searches tweets from @naval containing "life advice." If you liked it, you can spot it in the results by the heart icon being filled in.

The limitation:

This only works if you remember substantial context, the account, approximate date range, or unique phrases. It's detective work, not a search solution.

Best for: Tweets from accounts you frequently follow where you remember specific terminology.

Method 3: Download Your X Data Archive

Difficulty: Medium | Time Investment: High initial setup, low maintenance | Reliability: High

This is where things get more technical but significantly more powerful. X allows you to download your complete data archive, which includes every tweet you've ever liked in a structured format.

How to download your archive:

1. Go to Settings → Your Account → Download an archive of your data
2. Request your archive (X will email you when it's ready, usually about 24 hours, sometimes up to 72)
3. Download the ZIP file (can be several hundred MB)
4. Extract and open Your archive.html in a browser

What you'll find:

The archive contains a complete HTML viewer with:

• All tweets you've posted
• All tweets you've liked (with the tweet ID, its full text, and a URL, but no author and no record of when you liked it)
• Your DMs, followers, following list, and more

Searching within the archive:

The archive HTML includes a basic search function (Ctrl+F in your browser), but it's limited:

• Only searches visible text on the current page
• Doesn't search by meaning or concept
• No categorization or filtering
• Manual and tedious for large collections

Taking it further:

The real power comes from extracting the data files in the archive:

• like.js - Contains all your liked tweets. It is JSON wrapped in a window.YTD.like.part0 = assignment, so strip that prefix before parsing
• tweets.js - Your posted tweets (tweet.js in older archives)
• profile.js - Account metadata

Tech-savvy users can write scripts to parse this JSON and build custom search interfaces. For example, you could load like.js into a spreadsheet program, convert it to CSV, and use Excel/Google Sheets filtering.

Pros:

• Complete historical data
• Offline access
• Machine-readable format for custom tools
• One-time download covers all history

Cons:

• Archive wait of about 24 hours (sometimes up to 72)
• Requires technical skill to use effectively
• Basic HTML viewer has poor search functionality
• No automatic updates (need to re-download for new likes)

Best for: Power users comfortable with JSON/data files, or those wanting to build custom solutions.

Method 4: AI-Powered Knowledge Base Tools

Difficulty: Easy | Time Investment: Low | Reliability: Very High

This is the emerging solution category that treats your liked tweets as what they actually are: a personal knowledge base that deserves sophisticated search and organization.

How these tools work:

1. Upload your X data archive
2. AI processes every liked tweet to:
   • Create semantic embeddings (meaning-based fingerprints)
   • Classify by topic and knowledge domain
   • Extract key concepts and tags
   • Generate searchable metadata
3. Search by meaning, not just keywords
4. Browse by category, date, or custom filters

Real-world example:

Instead of searching for exact keywords like "productivity tips," you can search for:

• "How do successful people manage their time"
• "Methods to focus better at work"
• "Reducing distractions while working"

The AI understands conceptual similarity and surfaces relevant tweets even if they don't contain your exact search terms.

Why this matters:

Over months or years, you've likely saved tweets about:

• Career advice and business insights
• Technical tutorials and code snippets
• Book recommendations and reading notes
• Creative inspiration and design examples
• Philosophy and thought-provoking ideas

You curated this collection for a reason, but without proper search, it's trapped in chronological limbo. Tools in this category turn that raw data into an actual searchable, AI-organized reference library.

Example tool: X Brain ($4.99 one-time) uploads your archive and uses AI to classify tweets into 15 knowledge domains and 65+ subcategories, with semantic search and automatic tagging. You can search by meaning and export enriched data as CSV, JSON, Markdown, or an Obsidian-ready vault.

Pros:

• Searches by meaning, not just keywords
• Automatic AI categorization and tagging
• Handles thousands of tweets effortlessly
• One-time processing, persistent searchability
• Export capabilities for future use

Cons:

• Requires a one-time payment ($4.99 for X Brain; most alternatives are monthly subscriptions)
• Still needs X archive download first
• Dependent on third-party service
• May require API keys for some tools

Best for: Anyone with 500+ liked tweets who regularly needs to reference old content. Essential for researchers, content creators, and knowledge workers.

Method 5: Bookmark Tweets You'll Need Again

Difficulty: Easy | Time Investment: Low ongoing | Reliability: High (for future tweets only)

This isn't a solution for searching past likes, but it's a proactive habit that makes future tweets easier to find.

How to do it:

1. When you like a tweet you'll want to find again, also tap the bookmark icon
2. Open Bookmarks from your profile menu when you need it
3. Use the search box in Bookmarks, where your account and app version have it (X has been adding bookmark search since 2024)

Why bookmarks work better than likes for reference:

• Intentional curation: You're consciously deciding this is reference material
• Fully private: Nobody sees your bookmarks, not even the post's author (authors can still see who liked their post)
• Separate from casual likes: Your likes can remain for engagement/appreciation
• Searchable in the app: Bookmarks can be searched where X offers it; likes cannot

Limitations: Doesn't help with your existing liked tweets, only future ones. And bookmarks don't come back in usable form from your X archive (it holds tweet IDs, not the text), so they live only inside X.

(An earlier version of this guide suggested saving tweets to a private List. X Lists hold accounts, not individual tweets, so that doesn't work.)

Best for: Building a findable reference habit going forward while using other methods for historical tweets.

Comparison: Which Method Should You Use?

The Bigger Picture: Why This Problem Exists

X's design philosophy treats likes as ephemeral engagement signals, not as bookmarks for future reference. The platform earns money by keeping you in the real-time feed, not helping you resurface old content.

This is why X added a separate Bookmarks feature, which has been gaining a basic search box. But it only covers tweets you remembered to bookmark, and it doesn't help with the years of likes you already have.

For power users, the reality is clear: you need a system outside X's native features if you're serious about treating liked/bookmarked tweets as reference material.

Practical Workflow: How to Actually Solve This

Here's a recommended workflow that combines multiple methods:

Step 1: Download your X archive today (even if you don't need it yet)

• It takes about 24 hours (sometimes up to 72), so start the request now
• This gives you a complete backup of your like history

Step 2: Process your archive with an AI tool if you have 500+ likes

• One-time cost, permanent searchability
• Especially valuable if you use likes for professional reference

Step 3: Bookmark future important tweets as well as liking them

• Stop relying on likes alone
• Develop a habit of bookmarking reference material

Step 4: Set a quarterly reminder to re-download your archive

• Keep your searchable database updated
• Re-process with your tool of choice

Key Takeaway

The native X experience will never solve this problem well because it conflicts with the platform's business model. If you're a casual user who likes 5-10 tweets per week, manual scrolling might suffice. But if you're a researcher, content creator, or knowledge worker who's curated thousands of liked tweets over the years, you need a dedicated solution.

Your liked tweets represent hundreds of hours of curation, ideas, resources, and references you found valuable enough to save. Don't let that investment disappear into an unsearchable void.

---

Ready to turn your X likes into a searchable knowledge base? The first step is downloading your X data archive. Go to Settings → Your Account → Download archive, and while you're waiting for it to process, decide whether you want to tackle the JSON files yourself or let AI handle the heavy lifting. Either way, your future self will thank you when you instantly find that perfect tweet you saved eight months ago.

Frequently Asked Questions

How do I find a tweet I liked months ago?

X's native search doesn't cover old likes. Your options are: scrolling through your Likes tab (slow), using your Data Archive with ctrl+F (limited), or importing your likes into a dedicated tool like X Brain that provides full-text semantic search across your entire like history.

Does Twitter let you search your liked tweets?

Twitter/X has no built-in search for liked tweets. The Likes tab only lets you scroll chronologically with no keyword search. For searchable access to your liked tweets, you need to either download your data archive or use a third-party tool that indexes your likes.

What's the fastest way to search old Twitter likes?

Import your Twitter data archive into a semantic search tool like X Brain. This indexes all your likes with AI classification, enabling instant keyword and topic searches across thousands of tweets. Author filtering is not possible: the archive's like.js stores only the post ID, text, and URL, with no author field. Setup takes 5-10 minutes versus endless manual scrolling.

---

Related Articles

• Search Twitter Likes by Keyword, Username & Date (2026)
• How to Export & Backup All Your Twitter Likes (2026 Guide)
• Turn Twitter Bookmarks Into Searchable Knowledge Base