How Many Hashtags to Use on Each Platform (Instagram, TikTok, X)
A platform-by-platform hashtag guide: how many to use on Instagram, TikTok and X, relevance vs. trend size, how to pick them, and how to avoid getting throttled.
How Many Hashtags to Use on Each Platform (Instagram, TikTok, X)
The single most common hashtag mistake I see is treating every platform the same: copy the same 15-tag wall, paste it on Instagram, TikTok, and X, and wonder why two of the three posts flop. They flop because the algorithms behind those three apps disagree about what a hashtag even is. One rewards saturation, one rewards relevance, and one quietly punishes you for showing up with more than two. Getting the count right per platform is the cheapest reach you'll ever buy.
This guide covers the numbers that actually come from public studies, how to choose between a relevant tag and a trending one, and the throttling traps that make good tags backfire.
The real per-platform counts (with sources)
Here's the part most "use 30 hashtags!" listicles skip: the right number is different on every app, and there's data behind each one.
- Instagram: fill all 30. Instagram's hard cap is 30 tags per post, and Later's 2024 analysis of 18 million posts found that posts using all 30 hashtags earned +13% reach compared to posts using only 5. Instagram is the one platform where saturation genuinely pays — within its own limit.
- TikTok: 10–15. TikTok's algorithm weighs relevance far more than raw count. TikTok Creative Center's own guidance lands around 3–5 niche tags plus 2–3 trending tags like
#fypor#foryou, which caps a healthy post at roughly 10–15. - X (Twitter): 2–3, and no more. This is the counterintuitive one. A 2023 Buffer study of 2 million tweets found tweets with 1–2 hashtags got +21% engagement, while tweets with 3 or more dropped -17%. On X, a third hashtag is actively costing you.
Two adjacent platforms worth knowing if you post in Chinese: Xiaohongshu (小红书) likes 5–10 long-tail topic tags, and Weibo uses the literal #话题# wrapper syntax rather than a leading #. Those formats don't interchange, which is its own trap (more below).
Relevance vs. trend size: pick the right mix
A hashtag has two properties that pull in opposite directions: relevance (does it describe your post?) and trend size (how many people are searching it?). Big trending tags like #fitness reach a huge audience but bury you instantly — that tag sits under 500 million-plus posts, so your clip is gone in seconds. Tiny hyper-specific tags are findable but reach almost nobody.
The fix is a ladder, not a single rung. For a typical post I aim for roughly:
- 2–3 broad tags for ceiling reach (
#fitnessmotivation) - 5–8 mid-size niche tags that you can realistically rank inside (
#dumbbellworkout) - 1–2 platform-native trend tags where they fit (
#fypon TikTok) - your own topic-as-tag, always — it's the safest, most relevant tag you own
That last one matters more than it sounds. Your literal topic is the one hashtag guaranteed to be relevant and guaranteed not to be on any ban list.
A worked example: one reel, five platforms
Say you filmed a 20-second latte-art reel. Type latte art coffee shop cozy into the Hashtag Generator and here's the shape of what comes back, native to each app:
- Instagram (30):
#latteart #coffeeshop #cozyvibes #coffeelover #specialtycoffee #baristalife …(filled to the cap) - TikTok (12):
#latteart #coffeetok #fyp #foryou #cafevibes #baristalife … - X (2):
#latteart #coffee - Xiaohongshu (8):
#拿铁拉花 #咖啡日常 #探店 #咖啡馆 … - Weibo:
#拿铁拉花# #咖啡#
Same input, five different counts and two different syntaxes. The matcher scores 14 niches in parallel — a "cozy coffee" prompt blends the food and lifestyle pools instead of guessing one — and the score shown next to each niche lets you sanity-check that it actually locked onto "food at 9 points," not a random guess. If you want help shaping the caption and title that ride alongside those tags, the YouTube Title Analyzer scores headline strength on the same page-load-fast, client-side basis.
How to actually pick (and avoid getting throttled)
Choosing tags is half the job; not getting throttled is the other half. A few rules that have saved my accounts:
- Never paste the Instagram wall onto X. Thirty tags on a tweet isn't ambitious, it's a -17% engagement penalty. Use the platform-correct count every time.
- Don't copy the Weibo block elsewhere. The
#话题#wrapper only parses on Weibo. Drop the trailing#for any other platform, or just copy from that platform's own card. - Spot-check sensitive accounts. Instagram's shadow-ban list shifts weekly and is never officially published. A good generator already excludes the known broken-hashtag list (tags temporarily action-blocked for spam), but for a high-stakes campaign, search two or three of your tags inside Instagram first.
- Skip bot-magnet tags. Tags that mostly attract bots inflate impressions without inflating real engagement, which can train the algorithm to deprioritize you.
A trustworthy tool helps here by sourcing tags from public trending data — Instagram leaderboards, TikTok Creative Center discovery, Xiaohongshu explore pages, Weibo's 热门话题榜 — rather than inventing plausible-looking strings an LLM dreamed up. Invented tags are the fastest route to a dead post.
My own workflow
I save my winners. When my "autumn travel Kyoto" post hit 40k views, I stored that exact pack — topic, platform, tags, timestamp — in the favorites drawer. A month later I shot a similar trip, opened the drawer newest-first, and one-clicked the same set. No re-running the matcher, no trying to remember which 30 tags worked in October. The drawer caps at 50 entries in localStorage, so a full season of proven packs stays one click away, per browser, never uploaded. That recycling habit has been worth more to my reach than any single "viral tag" ever was.
Hashtags aren't a magic switch, but the count-per-platform discipline is close to free reach, and it's the part most people get wrong. Match the number to the algorithm, ladder your relevance against trend size, keep your topic-as-tag as the anchor, and recycle what works.
Made by Toolora · Updated 2026-06-13