WEIR
Get a key
Language中文日本語
Theme

Find Viral Tweets in Your Niche: min_faves, Top, Dates

Find viral tweets in your niche with min_faves, the Top tab and a date window, plus what 10 viral posts we pulled had in common and a Python script.


Published · Updated · 6 min read

Cover image: Viral this month, by brand

To find viral tweets in your niche, search X for your topic with min_faves:1000, choose the Top tab, and add a date window.

This guide is for creators, marketers and founders who want to see what works in their corner of X (Twitter). You get the searches, how to read the results, and what we learned from 15 viral posts we pulled on 26 October 2026. We run Weir, the API in Step 5.

Paste this into X's search box, swap MCP for your niche, and stay on Top:

MCP min_faves:1000 since:2026-09-26 lang:en -filter:nativeretweets
  1. Search with min_favesYour niche word plus a like floor.
  2. Pick Top or LatestBiggest first, or newest first.
  3. Add a date windowsince: and until: by day.
  4. Read what they shareFormat, media, bookmarks.
  5. Run it on a scheduleSame search by API or agent.

Step 1: Find viral tweets in your niche with min_faves

Two operators set the bar for "viral." min_faves: keeps posts with at least that many likes, and min_retweets: does the same for reposts.

Start with your niche word and a floor that fits its size:

  • Big niche (AI, crypto, sports): min_faves:5000.
  • Mid niche (dev tools, SaaS): min_faves:1000.
  • Small niche (a hobby, a city): min_faves:100.

Too many results? Raise the floor. Nothing back? Lower it by ten times.

Add -filter:nativeretweets to drop reposts, and -filter:replies to keep original posts only. Our X advanced search builder writes these for you.

Step 2: Choose Top or Latest

X's search has a Top tab and a Latest tab, and they answer different questions.

  • Top shows X's own pick of the most engaging posts. Use it to study what already won.
  • Latest shows matches newest first. With a like floor, it shows posts taking off this week.

Neither one sorts by likes. Here is what Top gave us for the search at the top of this page:

10posts on Top with 1,000+ likes
9thwhere the most-liked post sat
857likes on one post under the floor

So read every result and rank them by the count under each post. X's filter is loose too: a post a little under the floor can slip in.

Step 3: Add a date window

Without dates, Top fills with old hits. Add since: and until: with a date in YYYY-MM-DD form:

MCP min_faves:1000 since:2026-10-01 until:2026-10-25

until: includes its own day, so this covers 1 to 25 October. A short window needs a lower floor, because fewer posts clear it.

For a date range explained in depth, see how to search tweets by date.

Step 4: Read what the viral posts have in common

Finding viral posts is half the job. The other half is seeing what they share.

We read the 10 posts from our MCP search. All were from creators, so we report patterns, not names:

  • Every one had more bookmarks than likes. The top post had 9,110 bookmarks and 6,339 likes.
  • Nine of ten carried a photo or video. The one without was a product launch.
  • Six shared an open-source repo or tool. Lists like "15 GitHub repos for..." did well.
  • Two hyped a free one-hour course, each from a different account.

People save posts they plan to use later. In this niche, "save this" content beat opinion.

Then we checked six brands in the same space with fromUsers. Their top five posts since 26 September:

Brand posts with 2,000+ likes, 26 Sept to 26 Oct 2026
Anthropic: formal verification9,669
OpenAI: legacy languages9,564
OpenAI: Brotli security flaw6,884
Cursor: computer use6,377
Figma: Dev Mode in Codex2,351

Brand posts ran the other way. All five had fewer bookmarks than likes, so news gets liked and tutorials get saved.

Step 5: Run the same search through Weir

Weir's search takes the same filters as JSON fields: query, minLikes, since and sort. Through Claude, ChatGPT or Cursor with Weir connected, paste:

Search X for MCP posts with 1,000+ likes since 26 September 2026, top 10, English.
Sort them by likes and tell me what formats they share. No usernames.

In a script, this Python uses only the standard library, so it runs in agent sandboxes with no pip:

import json, os, urllib.request

def viral(query, min_likes, since, max_pages=3):
    body = {"query": query, "minLikes": min_likes, "since": since, "sort": "top", "limit": 20}
    headers = {"Authorization": f"Bearer {os.environ['WEIR_API_KEY']}", "Content-Type": "application/json"}
    posts = []
    for _ in range(max_pages):
        req = urllib.request.Request("https://api.weirapp.io/v1/x/search", json.dumps(body).encode(), headers)
        page = json.load(urllib.request.urlopen(req, timeout=60))
        posts += [p for p in page["data"] if (p["likeCount"] or 0) >= min_likes]
        if not page["meta"]["nextCursor"]:
            break
        body["cursor"] = page["meta"]["nextCursor"]
    return sorted(posts, key=lambda p: p["likeCount"], reverse=True)

Call it with your niche word, floor and start time:

for p in viral("MCP lang:en -filter:nativeretweets", 1000, "2026-09-26T00:00:00Z"):
    print(p["likeCount"], p["bookmarkCount"], p["url"])

It keeps only posts that really clear the floor, then sorts by likes. Weir's since field takes an exact time with its zone, while since: in X's box works by whole days.

The brand post that led our chart, trimmed:

{
  "url": "https://x.com/AnthropicAI/status/2113390212911083894",
  "text": "New on our Frontier Red Team blog: using Claude to formally verify software at large scale. ...",
  "createdAt": "2026-10-22T22:01:06.000Z",
  "likeCount": 9669,
  "bookmarkCount": 6590
}

What does it cost to track viral posts?

A search costs 1 credit plus 1 per post. Our two searches pulled 15 posts for 17 credits.

The script above reads up to 60 posts, or 63 credits, per run. Run it once a week and a month costs about 252 credits.

Weir's first 5,000 credits each month are free, with no card. Our X MCP server page lists every tool an agent can call.

If your viral search finds nothing

  • Your floor is too high. Drop min_faves: by ten times and try again.
  • Your window is too short. Widen since: to 30 or 90 days.
  • Your niche word is too rare. Try the word people use, or a hashtag.
  • Results are all reposts. Add -filter:nativeretweets; a repost shows 0 likes of its own.

When not to use Weir

  • You check one niche by hand now and then. X's search box is free.
  • You need posts from protected accounts. Weir reads public posts only.
  • You want to schedule or post content. Weir is read-only.
  • Price per post is all that matters. TwitterAPI.io lists $0.15 per 1,000 tweets as of October 2026, less than our smaller packs.

FAQ

How do I find viral tweets in my niche?

Search your niche word with min_faves:1000 on X's Top tab, then add since: for a recent window. Adjust the number until a page of posts remains, and compare the like counts yourself, because Top isn't sorted by likes.

What counts as a viral tweet?

There is no fixed number. It depends on the niche's size. In our MCP test, 1,000 likes over one month still filled a full page, so big niches need 5,000 or more.

Can I find viral tweets from last week only?

Yes. Add since: with the date a week ago, such as since:2026-10-19, and stay on Top. Switch to Latest to see posts still climbing.

Sources

All checked on 26 October 2026:

Keep reading