AI Content Creator Prompts — 14 Copy-Paste Templates for YouTube, Newsletter, Threads & Repurposing (May 2026)
Free prompt enhancer: midastools.co/prompt-enhancer — paste a rough hook or headline, get a tighter, attention-grabbing version back in seconds. Deep dive: The 1-to-10 AI Content Repurposing Workflow (May 2026)
The top 1% of creators don't out-write anyone. They out-leverage. One blog post becomes 6 platform-specific posts. One YouTube video becomes 12 short-form clips, 4 Twitter threads, 2 newsletter sections, and 30 days of social content. The math compounds — and almost all of it can be drafted in AI, refined in 5 minutes per asset, and shipped.
This is the working set I keep pinned: 14 prompts tested against real publishing workflows for YouTube creators, Substack/Beehiiv writers, threadboi-level Twitter accounts, and LinkedIn solo brands in the last 30 days. Copy, paste, fill the brackets, ship.
| # | Template | When to use |
|---|---|---|
| 1 | Master Multi-Platform Repurposer | One blog/video → 6 platform-tailored posts at once |
| 2 | Twitter Thread from a Blog Post | Turn a 1,200-word essay into a 9-tweet thread |
| 3 | YouTube Hook (first 30 seconds) | The single most important 30 seconds in your video |
| 4 | YouTube Long-Form Script (10–15 min) | Full script with hook + sections + CTA |
| 5 | YouTube Shorts Script (under 60 sec) | Vertical-first script with pattern-interrupt opener |
| 6 | Newsletter Weekly Roundup | Substack/Beehiiv format with 3-5 curated items |
| 7 | Newsletter Deep Dive (one big idea) | When you have something serious to say |
| 8 | LinkedIn Carousel (8 slides) | Highest-engagement format on LinkedIn in 2026 |
| 9 | Twitter Hot Take | Single tweet that earns the reply guys' attention |
| 10 | Twitter Story Thread | Personal narrative thread (5–7 beats) |
| 11 | Curiosity-Gap Headline Generator | 5 title variants that don't feel like clickbait |
| 12 | Benefit-Driven Subject Line Generator | 5 newsletter subject lines optimized for open rate |
| 13 | 30-Day Content Calendar | Theme + pillar + daily post type for a full month |
| 14 | Content Pillar Generator | The 4–6 topics your account should own |
Every content prompt worth reusing has seven slots. Fill them in order:
[ROLE]: who the AI is writing as (e.g. "senior content strategist for a B2B SaaS")
[PERSONA]: your audience — who reads/watches (their job, level, what they care about)
[PLATFORM]: the destination (Twitter, LinkedIn carousel, Substack, YouTube Shorts)
[SOURCE]: the raw material to work from (blog draft, video transcript, idea)
[GOAL]: what success looks like (newsletter signups, replies, shares, click-throughs)
[VOICE]: your tone — pick ONE (warm-direct / contrarian / earnest / playful / academic)
[CONSTRAINTS]: format, length, what to avoid, what to include
Skip any slot and the model fills it with filler — "in today's fast-paced world," "are you ready to take your content to the next level?", "the key is consistency." The exact phrases your audience has trained themselves to scroll past.
The highest-ROI prompt in this kit. One asset → six platform-specific versions in one shot.
You are a senior content strategist for an audience of [PERSONA].
Here is my source content:
"[PASTE BLOG POST / TRANSCRIPT / DRAFT]"
Turn it into six platform-tailored versions:
1. TWITTER/X THREAD — 8 tweets max, each under 270 chars, lead with the
strongest insight, hook in tweet 1, single CTA at end.
2. LINKEDIN POST — 1,300 chars, conversational opener (first 2 lines hook
before "see more"), one personal beat, one numbered list of 3 takeaways,
end with a question.
3. INSTAGRAM CAPTION — 200 words max, first line is the hook, line break
every 1–2 sentences, ends with 5–8 hashtags relevant to [PERSONA].
4. YOUTUBE SHORTS SCRIPT — 45 seconds, vertical, pattern-interrupt opener,
3 fast beats, hard CTA frame at the end.
5. NEWSLETTER SECTION — 250 words, subject-line ready, one specific example,
one actionable takeaway.
6. THREADS / BLUESKY POST — 500 chars, less formal than Twitter, conversational,
inviting one reply not retweets.
Constraints:
- Maintain the source's actual claims; do not invent statistics.
- Match my voice: [VOICE].
- Goal across all 6: [GOAL].
Try refining each output with the free prompt enhancer before publishing — most "I'll just post it" drafts have a flabby line that bleeds engagement.
You are a thread-writer for an audience of [PERSONA].
Source blog post:
"[PASTE FULL POST]"
Turn it into a 9-tweet thread.
Constraints per tweet:
- Tweet 1 (HOOK): a single specific claim or contrarian observation. No "🧵"
emoji. No "let me explain". Just the claim. Under 270 chars.
- Tweets 2–7: one beat per tweet. Each tweet must be readable standalone.
Use line breaks liberally. No tweet starts with "and".
- Tweet 8 (PAYOFF): the actionable takeaway readers can use today.
- Tweet 9 (CTA): one line about the source post + link. No bait language.
Avoid:
- "1/", "2/" numbering (Twitter shows position natively)
- Em dashes (they read as AI)
- Bullet lists inside a tweet
- Generic engagement asks ("agree?")
Hook strength check: tweet 1 should make a reader stop scrolling. If it
sounds like a Medium headline, rewrite.
The single most important 30 seconds in your video. AVD drops off a cliff if the first 30 seconds don't earn the next 60.
You are a YouTube retention specialist. My video is about [TOPIC] for
[PERSONA]. The full video is [LENGTH] long.
Write 3 hook variants for the opening 30 seconds. Each must:
1. Start with a pattern interrupt — surprising claim, weird visual cue,
or a question that makes the viewer say "wait, what?"
2. Within 8 seconds: state the specific outcome they get by watching
to the end. (Not "we'll explore" — the actual deliverable.)
3. Within 20 seconds: name the obstacle / mistake most people make
that this video solves.
4. Within 30 seconds: transition into the body with a clear "here's
exactly what we're doing" beat.
Constraints:
- Spoken language (write for ear, not eye). Contractions on.
- Max 80 words per hook (~30 seconds at conversational pace).
- No "in this video, we'll cover" — show, don't tell.
- No "smash that subscribe button" — it kills retention.
Output 3 distinct hook directions, not 3 variants of the same hook.
You are a YouTube script writer. Topic: [TOPIC]. Audience: [PERSONA].
Length target: 12 minutes (~1,800 spoken words). Goal: [GOAL] (e.g.,
viewer subscribes, clicks a specific link in description).
Structure the script:
1. HOOK (0:00–0:30) — pattern interrupt + specific outcome + obstacle.
2. CREDIBILITY BEAT (0:30–1:00) — one sentence on why I'm the one
teaching this. Concrete, not "I've helped thousands."
3. ROADMAP (1:00–1:30) — name the 3 sections we're covering.
4. SECTION 1 (1:30–4:30) — first section with one specific example.
5. RE-HOOK (4:30–4:45) — "if you're still with me, here's the part
most people get wrong" or similar.
6. SECTION 2 (4:45–8:00) — second section, contrasting angle.
7. SECTION 3 (8:00–11:00) — third section, the actionable payoff.
8. RECAP + CTA (11:00–12:00) — 3-bullet recap, single CTA, sign-off.
Write the full script as it would be spoken — contractions, short
sentences, occasional sentence fragments. Add (B-ROLL: …) cues where
you'd cut to footage. Add (PAUSE) where to pause for emphasis.
Voice: [VOICE]. Avoid: jargon, "as I mentioned earlier", saying
"obviously" or "literally" as filler.
You are a Shorts script writer. Topic: [TOPIC]. Audience: [PERSONA].
Constraints:
- Total length: 45–55 seconds spoken (~120 words).
- Vertical mindset: punchy, fast, one idea.
- First 3 seconds: a hard pattern interrupt — either a contrarian claim
("Stop posting daily. Here's why."), a visual demo cue, or a specific
number ("I shipped 47 videos last year. Here's the one that 10x'd my
channel.").
- Beat 2 (3–15s): the specific insight or method, one sentence.
- Beat 3 (15–35s): the proof — example, before/after, specific result.
- Beat 4 (35–50s): the actionable takeaway viewers can use today.
- Beat 5 (50–55s): single CTA — comment a keyword, follow for part 2,
or click bio link. Pick ONE.
Voice: [VOICE]. Avoid asking "have you ever felt…" openers, fake
urgency, and "if you got value from this, like and subscribe" boilerplate.
You are writing a weekly roundup newsletter for [PERSONA].
This week's items to cover:
- [ITEM 1: title + 1-line context + link]
- [ITEM 2: title + 1-line context + link]
- [ITEM 3: title + 1-line context + link]
- [ITEM 4: title + 1-line context + link]
- [ITEM 5: title + 1-line context + link]
Format:
- Subject line: under 50 chars, curiosity-driven, no "Weekly digest" framing.
- Preview text (preheader): 80 chars that pay off the subject line.
- Opener: 2-sentence personal note from me. Set the week's frame.
- For each item: title (link), 2-sentence why-you-should-read, no fluff.
- Closing: one open question that invites a reply.
- Signature: [MY NAME]
Voice: [VOICE]. Length: 350 words max body. Reading time target: 90 seconds.
No "happy [day-of-week]", no "I hope you're well". Skip pleasantries; lead
with substance.
You are writing a deep-dive newsletter for [PERSONA]. One big idea, one
issue.
The idea: "[STATE YOUR THESIS IN ONE SENTENCE]"
Structure:
1. SUBJECT LINE — curiosity, 50 chars max. Avoid "5 ways to" formats.
2. PREHEADER — pays off the subject. 80 chars.
3. HOOK PARAGRAPH (60 words) — concrete scene, specific example, or
contrarian observation that introduces the idea.
4. WHY THIS MATTERS NOW (80 words) — what changed in the last 12 months
that makes this thesis relevant now (not 2 years ago).
5. THE CORE ARGUMENT (300 words, 3 sub-beats):
- Beat 1: what most people believe (and why)
- Beat 2: where that belief breaks down (one concrete example)
- Beat 3: the better mental model
6. WHAT TO DO ABOUT IT (100 words) — 3 specific actions a reader can
take this week. Not "think about it" — actions.
7. CLOSER (40 words) — one sentence personal beat + one open question.
Voice: [VOICE]. Avoid academic jargon, "we'll explore", and frameworks
named after acronyms unless the framework is yours.
You are a LinkedIn carousel writer for [PERSONA]. Topic: [TOPIC]. Goal:
[GOAL] (e.g., comments, profile visits, DMs).
Output 8 slides with text only (designer will style):
SLIDE 1 (HOOK COVER):
- Top line: contrarian claim or specific number that stops the scroll.
- Bottom line: who this is for. 1 sentence.
SLIDES 2–7 (one idea per slide):
- Top: section heading (5 words max).
- Body: 2–3 short lines. No paragraphs.
- Optional: one concrete example.
SLIDE 8 (CTA):
- Top line: one-sentence summary of the whole carousel.
- Body: ONE specific action — "Follow for next week's breakdown" / "Save
this for your next [SITUATION]" / "Comment [WORD] for the template."
- Avoid: "What did I miss?" / "Agree?" / generic "thoughts?"
Constraints:
- Max 25 words per slide (excluding cover and CTA).
- Voice: [VOICE].
- Avoid AI tells: em dashes, "delve", "leverage", "in today's landscape",
"navigate the complexities of".
You are me. Write a single tweet that takes a real position.
Topic: [TOPIC]. My actual belief: "[YOUR ACTUAL BELIEF]"
Constraints:
- Under 270 chars (room for one quote-tweet).
- State the position clearly in the first 15 words.
- Optional: one specific example or number after the position.
- Optional: a brief "because…" beat (max 15 words).
- No "unpopular opinion:" prefix (it's become a tell).
- No "hot take:" prefix (same).
- No "controversial but…" (same).
- No mealy hedges ("kinda", "maybe", "tbh").
Goal: a position the reply guys will engage with. Either they agree
hard or they disagree hard — both are fine. Lukewarm is the loss state.
You are me, writing a personal narrative thread for [PERSONA].
The story: "[YOUR STORY — what happened, who was involved, what changed]"
Structure (5–7 tweets):
TWEET 1 (HOOK):
- The outcome first (the surprising result).
- 1–2 sentences.
- Don't tease — give the punchline. The story is the proof.
TWEETS 2–5 (THE BEATS):
- One beat per tweet. Show, don't summarize.
- Use specifics: real numbers, real names (if appropriate), real dialogue.
- Each tweet readable standalone.
TWEET 6 (THE TURN):
- The moment something changed. The decision, the realization, the pivot.
TWEET 7 (THE LESSON):
- What this means for the reader. One actionable takeaway.
- Single CTA if relevant. No "If you found this useful, RT" begging.
Voice: [VOICE]. Avoid: starting tweets with "Then", "So", or "And".
Avoid wrapping with "Anyway, that's my story" — let the lesson tweet
do the work.
You are a headline writer. The piece is about [TOPIC]. Audience: [PERSONA].
Platform: [BLOG / NEWSLETTER / YOUTUBE].
Generate 5 headline variants. Each must:
1. Use a curiosity gap (something the reader doesn't know yet but wants to)
2. Be specific (numbers, named entities, concrete outcomes)
3. Pass the "would I actually click this" test
4. Avoid clickbait tells: "you won't believe", "what happened next", "doctors hate",
"this one weird trick"
Output format:
1. [HEADLINE]
Why it works: [1-line explanation of the curiosity mechanism used]
Variants should use 5 DIFFERENT curiosity mechanisms (not 5 of the same):
- The unexpected result ("I tried X for 30 days. Here's the surprise.")
- The contrarian frame ("Why most advice about [TOPIC] is wrong")
- The specific number ("The 3 mistakes I made before [OUTCOME]")
- The named entity comparison ("[BRAND A] vs [BRAND B] — the honest review")
- The promise-with-cost ("[OUTCOME] in 30 minutes — but here's the trade-off")
You are an email subject line writer for a newsletter to [PERSONA]. Topic
of this issue: [TOPIC].
Generate 5 subject lines optimized for open rate. Constraints:
- 30–50 characters each (mobile preview cutoff)
- Lead with the benefit or curiosity, not the topic name
- Test against the "open in 2 seconds" decision a reader makes in their inbox
- One variant should use a question
- One variant should use a specific number
- One variant should use a personal/conversational tone ("I tried X…")
- No "Newsletter #47" framing
- No emoji unless your brand voice already uses them consistently
- No ALL CAPS
Output format:
1. [SUBJECT LINE] ([CHARACTER COUNT])
Hook type: [curiosity / benefit / number / question / personal]
Why this might work: [1-line note on the mechanism]
Output 5 variants using 5 DIFFERENT mechanisms.
You are a content strategist for [PERSONA]. Build a 30-day publishing
calendar for [PLATFORM(S)].
Inputs:
- My niche: [NICHE]
- My positioning: "[1-SENTENCE POSITIONING]"
- My pillars: [3–5 CONTENT PILLARS]
- Posting frequency: [TIMES PER WEEK PER PLATFORM]
- Goal for the month: [GOAL — e.g., 500 newsletter signups, 100 demo calls,
10K followers]
Output a day-by-day calendar with:
Day | Platform | Pillar | Post Type | Working Title | Format
----|----------|--------|-----------|---------------|--------
Constraints:
- Distribute pillars roughly evenly across the month
- Mix post types: educational, story, opinion, list, behind-scenes, case study
- At least 2 "anchor" pieces (deep dives, signature posts) in the month
- Include 1 dedicated promotion post per week (product, lead magnet, CTA)
- Include 1 "evergreen repurpose from archive" slot per week (saves time)
- No two consecutive days should be the same post type on the same platform
After the table, add a 1-paragraph note explaining the strategic logic
of the month (the throughline, the build, what success looks like at day 30).
You are a positioning strategist. Help me identify the 4–6 content
pillars I should own.
My context:
- Niche: [NICHE]
- Audience: [PERSONA]
- My unique angle: "[WHAT YOU KNOW THAT OTHERS IN YOUR NICHE DON'T]"
- My business goal: [GOAL — what content needs to convert to]
- Topics I'm already known for (if any): [LIST]
Generate 4–6 content pillars. Each pillar should:
1. Have a 2–4 word label (the topic I "own")
2. Connect to my unique angle (not generic niche topics)
3. Have 3 example sub-topics or post angles
4. Have a "why this matters" 1-sentence frame for my audience
5. Have an indicator of how it ladders to my business goal
Constraints:
- Pillars should be distinct (no overlap)
- At least one pillar should be the "anchor" — the topic I want to be
known for above all
- At least one pillar should be a "trojan horse" — broad-appeal topic
that funnels into my deeper niche
- Avoid pillars that are generic ("productivity", "mindset", "business")
Output as 5 markdown sections with each pillar's name as the heading.
After the pillars, add 2 example posts that show how to combine 2
pillars in one piece — pillar combinations are how distinct voice emerges.
| Task | ChatGPT (4o/5) | Claude (Sonnet/Opus) | Gemini 2.5+ | Best Pick |
|---|---|---|---|---|
| YouTube long-form scripts | Solid | Best | Solid | Claude — handles narrative arc + voice consistency over 2K words |
| Twitter threads | Best | Very good | Good | ChatGPT — best at punchy, scroll-stopping rhythm |
| Newsletter deep dives | Good | Best | Good | Claude — nuance + sustained argument |
| LinkedIn carousels | Best | Very good | Good | ChatGPT — better at "platform native" phrasing |
| Subject lines / headlines (variants) | Best | Good | Good | ChatGPT — better at variant generation |
| Repurposing (1 → 6 formats) | Very good | Best | Good | Claude — preserves source intent better |
| Content calendars | Good | Good | Best | Gemini — better at structured table output, Google ecosystem |
| Image prompts for thumbnails | Good | Good | Best | Gemini (Imagen) — closer to creators' workflow |
| YouTube hook brainstorming | Best | Very good | Good | ChatGPT — variant generation + tone control |
| Twitter hot takes | Very good | Best | Good | Claude — sharper on "actual position" framing |
The pattern: Claude for nuance, ChatGPT for variants, Gemini for structured outputs and image work. Most working creators end up using two of the three, not one.
-
Skipping the [PERSONA] slot. "Write a Twitter thread about productivity" returns generic LinkedIn-style sludge. "Write a Twitter thread about productivity for indie SaaS founders who already use Notion" returns something with edges. The model needs to know who the reader is, not just what the topic is.
-
Asking for "engaging" content. Engaging is what your audience decides, not what you brief. Replace "make it engaging" with specific constraints: "first line under 12 words", "no question in the first sentence", "use a specific named example."
-
One prompt, one output. The best creator prompts are CHAINS. Repurpose → expand → cut. Repurpose first, then run each output through prompt #11 (headline gen) or #12 (subject line gen). Layering 2–3 of these prompts beats any single super-prompt.
-
Not feeding the model your voice. Paste 2–3 paragraphs of your actual writing into the same chat thread BEFORE asking for content. The model picks up sentence rhythm, vocabulary, and tone way better from examples than from "write in a casual tone." Save 3 of your best paragraphs as a reusable "voice snippet."
-
Polishing the wrong asset. Most creators spend 30 minutes polishing the third paragraph and 30 seconds on the title. The hook (tweet 1, headline, subject line, first 30 seconds of a video) is where 80% of the work should go. Use prompt #11 / #12 / #3 on your best content even if you wrote the body yourself.
- Free prompt enhancer — paste a rough hook or headline, get a tighter, attention-grabbing version: midastools.co/prompt-enhancer
- 5 free prompts to try first — no signup, instant access: midastools.co/free-prompts
- Full Content Creator AI Kit — the complete 6-file workflow (repurposing engine + thread generator + YouTube script builder + newsletter writer + calendar generator + headline formulas). $39 one-time: midastools.co/content-creator-kit
- The 1-to-10 AI Content Repurposing Workflow — the deep dive blog post that walks through repurposing one piece of content into 10 formats: midastools.co/blog/ai-content-repurposing-2026
- Sister gist — AI Email Prompts (14 templates for replies, follow-ups & inbox zero) — the daily-tax companion to creator prompts.
Made by Midas Tools. All prompts here are tested against real creator workflows in 2026. Save this gist — you'll come back to it.