Twitter Competitor Analysis: AI Winning Tweets & Profile Posts

The tweets that beat each account's average, in your inbox weekly — with why they won.

⚡ First 25 winning tweets free · 💵 $0.003 per winning tweet · 🤖 ≈ $0.022 per winning tweet all in with AI, on plans with model access or your key — or ours · ⏱️ demo in seconds

Try it in 30 seconds. Click Try it — 54 sample tweets, their 10 winners in seconds, nothing charged. Then pick a one-click preset or type the X accounts you want to study. → What a run costs · What the AI adds · Run it weekly

⚡ At a glance

🎯 What this Actor does

You give it X accounts; it keeps and charges only the tweets that beat each account's normal:

• A fair winner test. Each tweet is measured against its own account's median, so a small account's breakout counts: "5.13× this account's median".
• Why it won. Rule tags (format, hook, length, hour) and, with AI, hook type, topic, CTA and tone.
• Content ideas for your brand. With AI and a one-line brand description: a remix idea and a draft post that reuse its structure, never its words.
• The pattern across winners. The hooks, formats and posting hours winners share ("5 of 10 are threads").
• Only new winners. Never charged twice; retweets, replies and ads are skipped free.

Find top performing tweets of any account

The multiple (default 2), window, min likes, formats, focusTopic, topPerAccount, excludeTerms and suppressionList are free filters.

Tweet hook analysis without the spreadsheet

The Why they won view lines up each winner's first line, hook pattern and multiple per account.

Twitter content ideas from competitors

Fill in "Your brand (for the remix drafts)" for a remix idea and a draft post per winner.

🚀 How to use it

1. Click Try it — the form is pre-filled with the free demo on sample tweets; it works as is.
2. Pick a preset or type your accounts, your topic and brand — that starts the live read. Leave "AI assessment" on: every winner gets a live AI verdict on any plan — Apify's model access, else our key (Keys / setup).
3. Press Start and open the dataset — winners land best-first; export to CSV or Excel, or wire a webhook.
4. Schedule it weekly (Actions → Schedule) with your address in "E-mail me the new winning tweets" — only new winners are delivered and charged.

📦 Already scraping X? Find the winners in your tweet scraper data

Pick your last run of Tweet Scraper V2, an X Tweet Scraper or a profile scraper in Dataset: no second scrape, you pay only for the winners.

📚 Learn more: Step-by-step tutorial — Twitter competitor analysis with this Actor

🆚 Why this instead of running the scraper yourself?

The scraper returns raw rows; the medians, the sorting and the "why" are left to you.

vs other X competitor analysis tools

More rivals (raw rows, no baseline or AI): scrape.badger/twitter-tweets-scraper $0.00015 per tweet · 605 users; scraper_one/x-posts-search $0.00025 · 508; scraper_one/x-profile-posts-scraper $0.0004 + $0.0025 per run · 457 (our 2nd source); viralanalyzer/twitter-scraper $0.005–$0.02 · 4; xtdata/twitter-x-user-tweets-scraper $0.0005–$0.0015 · 43, 50-tweet minimum; 6sigmag/twitter-profile-hunter-bulk-website-to-twitter-mapper $0.002 per profile · 0, websites → handles.

Raw-row cells assume our ≈ 12 % hit rate (830 tweets per 100 winners).

Store prices and users, checked 2026-10-17. None has a per-account baseline or an AI why-it-won.

💸 Cheaper than a Tweet Hunter or Hypefury seat

Tweet Hunter is $49 a month (Grow plan, checked 2026-10-17), Hypefury $19 (Full plan, checked 2026-10-17); 5 accounts weekly here ≈ $1.25 a month, ≈ $1.55 with AI.

💵 Pricing — what a run really costs

Pay per result: one event per winning tweet, plus one digest per run that billed a winner.

Never charged by us: losers, retweets, replies, ads, filtered or off-topic tweets, repeats, the demo. The X scan is not ours to waive — the scraper bills your account per tweet read, capped by "Max source spend"; a dataset you already have costs no scan.

How that compares — $0.003 is 1.3× the $0.0022 median of the 10 priced Actors next to this one in Store search for "twitter competitor analysis", a median of raw and follower scrapers. Against the 10 same-job X Actors (viral-tweet finders, post-performance monitors) it is under half their $0.007 median. One winning tweet replaces ≈ 8 raw rows, ≈ $0.0004 per raw row equivalent.

Worked example — 500 tweets read → ≈ 60 winners (≈ 12 %, estimate): X scan $0.25 on your account + 60 × $0.003 = $0.43, ≈ $0.007 per winning tweet with rules only. With AI: $0.25 + 60 × $0.015 + ≈ $0.18 tokens = $1.33, ≈ $0.022 per winning tweet. Your first 25 winners and first 3 digests are free, so your first run costs $0.25 + 35 × $0.003 ≈ $0.36.

At volume — an agency, 10 clients × 5 accounts weekly: $2.50 scan + 50 winners × $0.003 + 10 digests × $0.02 ≈ $2.85 a week (≈ $12.30 a month), ≈ $3.60 with AI — against one $49 Tweet Hunter seat.

AI tier, all in: ≈ $0.022 per winning tweet — $0.015 here + ≈ $0.003 tokens (1,725 in / 273 out, measured 2026-10-17, google:gemini-3.5-flash-lite) + ≈ $0.004 scan; on our key the tokens are inside the $0.015.

🤖 What the AI tier adds

The same winning tweet, AI off (the free demo run):

{"name": "@contoso_growth: 7 landing page lines that doubled our demo bookings: (sample — link inert)", "format": "thread",
 "outlierMultiple": 5.13, "ruleScore": 79, "label": "hot", "hookType": "number_or_list", "ctaType": "none",
 "whyItWorked": "Beat the account median by 5.13× with 640 likes and 520 bookmarks; thread, number opener", "remixIdea": null}

AI on ($0.015) — our 2026-10-17 run of the AI example input (on Apify's model access the default model is anthropic/claude-haiku-4.5):

{"name": "@contoso_growth: 7 landing page lines that doubled our demo bookings: (sample — link inert)", "fitScore": 94, "score": 94,
 "label": "hot", "assessedBy": "ai", "aiModel": "google:gemini-3.5-flash-lite",
 "hookType": "number_or_list", "contentPillar": "landing page copy", "ctaType": "none", "tone": "educational",
 "whyItWorked": "The tweet secured 520 bookmarks (an 0.81 bookmark-to-like ratio) and reached 5.13x the median by opening with a clear number hook: \"7 landing page lines that doubled our demo bookings:\" followed by actionable examples.",
 "draftPost": "7 onboarding steps that doubled our trial upgrades: → \"Invite your first teammate\" → \"No setup required\" → \"Import data in 1 click\" → \"Used by [number] founders\" → \"Test with your live metrics\" → \"Chat with an analyst\" → \"Free migration support\""}

The rule tier gives the pattern; only the AI names the mechanism and writes your version — about 5 minutes per tweet by hand (our estimate, not timed). Which plan gets which AI: Keys / setup.

⚙️ Input

📤 Output

One dataset item per winning tweet — export as JSON, CSV or Excel, or stream it to a webhook. Views: Shortlist,

Why they won

Output fields

Upsert key for a CRM or Sheet: dedupeKey — tw: plus the tweet id, so one tweet is one row however many runs or account lists find it; itemId is a hash of the same key.

A DIGEST (new and hot counts, what changed since the last run — "vs last run: qualified 41 → 55, hot 5 → 9" — the top 10 with links) and a client-ready REPORT page (OUTPUT.reportUrl) follow every run that delivered.

⭐ Found it useful? A review on the Store helps others find it.

🔁 Run it weekly

Track competitor tweets on autopilot:

1. Open the saved task and click Schedule after your first real run (its status links it). Put your address in "E-mail me the new winning tweets" and turn on "Fail the run when the source delivers nothing".
2. Leave "Never output the same winning tweet twice" on: each run keeps the 90-day median but delivers and charges only the winners posted since the last run.
3. Digest to Slack, Teams, Google Chat or Discord; rows to Sheets or your CRM via a Make / n8n / Zapier "create or update row" step on the webhook (tracker template, upsert on dedupeKey) — no weekly import; remix drafts to Typefully, Buffer or Hypefury the same way (scheduler recipe).

Rising now: preset Rising now on a schedule every 6 hours (0 */6 * * *) with your Slack URL in "Send alerts to" — the 30-day median stays, only the last day's winners arrive, while the reply window is still open.

Weekly cost: 5 accounts → $0.25 scan + ≈ 5 new winners × $0.003 + $0.02 digest ≈ $0.29 with rules, ≈ $0.36 with AI. OUTPUT.history, the REPORT page and historyDatasetName keep every run's winners.

🎯 Try it for your niche

Pick one in Start from a ready niche search and press Start, then schedule it. Each reads 3 accounts × 50 tweets, capped at $0.15 of source spend; its measured match count is dated 2026-10-24.

🔌 Integrations, automation and API

Which delivery field wins: each URL is posted once — use "Send alerts to"; the older per-service fields and their precedence are in the API reference.

• After every run of your X scraper (apidojo/tweet-scraper, kaitoeasyapi's, xquik's, the profile scraper): its Integrations tab → "Run succeeded → start this Actor" with sourceMode: dataset, datasetId = resource.defaultDatasetId, enableAi: false — rules only.
• Webhook — one POST per winner, or winning_tweets.qualified batches, each with runId; a 429 / 5xx is retried once and never fails the run.

Signature and which call runs what (OUTPUT.mode: {} → free demo · accounts → live, charged · list / dataset → your rows, no scan) are in the API reference.

Run outcomes — what your integration sees

👥 Who is it for?

🧠 How the AI works

The model sees only winners. Every why must quote the tweet and name a number; a remix copying 7 words of the original is dropped, handles and hashtags are stripped, an invented % or $ becomes [your number]. A model failure leaves the rule tags.

🔒 Data, compliance and limits

• Public posts of public X accounts only; follow X's terms and the law (GDPR, CCPA) for your use.
• An account needs 3 posts in the window for a median (under 8 is flagged low_baseline_confidence); the source's Free plan reads 50 tweets per run.

❓ FAQ

What will my first real search cost? 3 accounts × 90 days ≈ 300 tweets ≈ $0.15 on your account and ≈ 36 winners — the first 25 free, the rest $0.003 each with rules.

What are the top tweets by account worth copying? The ones far above that account's own median. The hook, format and posting hour that repeat across winners in the Why they won view are the answer for your niche.

Is this a Twitter profile posts scraper? Not a raw one — it runs the Twitter Profile Scraper on your account and adds the baseline, filter, analysis and memory.

Is it a viral tweets finder? Yes, per account: "viral" means far above that account's normal, so a small account's breakout counts as much as a big account's.

Is this a Tweet Hunter alternative? For competitor research, yes, at the seat prices above. It does not post for you — the drafts go to Typefully, Buffer or Hypefury as drafts (the scheduler recipe under Run it weekly).

List-mode rows

🧩 More Actors from the same developer

• Instagram Competitor Content — The same beat-the-median test for your competitors' Instagram posts.
• Facebook Page Post Tracker — The winning posts and offers of the Facebook pages you track, weekly.
• Threads Brand Monitoring — Every Threads post that mentions your brand or a competitor, with its sentiment, as an alert.
• Twitter Lead Generation — The X posts of people asking for a product like yours, ranked as leads with a reply draft.

Same competitors on every network: schedule this Actor, the Instagram and the Facebook one weekly, each webhook into one Sheet upserted on dedupeKey.

🆘 Support

Something missing or wrong? Open an issue on the Actor's page — buyer requests ship first.

📝 Changelog

• 0.1.3 (2026-10-18) — live AI verdicts in the free demo on every plan; drafts whose count disagrees with their list are fixed or dropped; Rising now preset.
• 0.1.2 (2026-10-18) — AI on every Apify plan at the same price; the big X scrapers' datasets read as they are; niche presets; a $0.02 digest per billed run (first 3 free).
• 0.1.1 (2026-10-17) — initial release: winners vs each account's median, rule tags, AI why-it-worked and remix draft, weekly digest.