Last updated: July 2026.
This AI influencer prompt library is 50 tested prompts, grouped into 10 scene types, and engineered around one job: keep the same character, voice, and framing consistent from post to post so followers actually recognise the account. Every prompt uses [BRACKETS] for the parts you fill in (character name, product, hook) and a fixed skeleton for the parts that never change: outfit style, lighting, shot type, delivery. That fixed skeleton is what makes a prompt library compound. Instead of rewriting a full prompt each morning, you paste the scene, swap two placeholders, and ship. If you have not designed the character yet, start with the character-save workflow; this library assumes you already have one saved.
Before you copy any prompt below, save your character. If your tool re-rolls the face on every render, the best prompt in the world will still produce a different person. The saved character is doing 70 percent of the consistency work; the prompt is doing the other 30.
Why a reusable prompt library compounds
Character consistency is not a per-prompt problem. It is a per-account problem. Followers stop scrolling because a face and a voice look and sound the same as the one they saw yesterday. That recognition takes 30 to 60 exposures to build, which means your account is running a marathon of near-identical renders, not a series of one-off creative experiments.
Rewriting the prompt from scratch each day fights that. Every subtle change (a new outfit adjective, a different lighting word, a new camera angle) nudges the render into slightly different territory. Ten posts in, the account reads as a lookalike, not the same creator.
A prompt library fixes it by locking the parts that matter and freeing the parts that do not. The lit-from-the-front, warm-tone, tight-frame skeleton stays the same across all 50 templates. What changes is the setting, the product, and the script hook. Followers see the same person doing different things, which is exactly how a real creator's feed works.
Specialist tools like Higgsfield and The Influencer AI both encourage prompt reuse for the same reason. The MakeAIVideo character pipeline is built around it: your saved character plus a stable prompt skeleton is the entire consistency system.
How to adapt these prompts
The prompts below are not meant to be run untouched. They are skeletons. Three quick rules for adapting them without breaking consistency:
Change the placeholders, keep the frame. Everything in [BRACKETS] is meant to be swapped: character name, product, location, script hook. Everything outside the brackets (shot type, lighting cue, delivery style) should stay put. That is what keeps 50 different scenes reading as the same creator.
Keep the script short. Each prompt ends with the character delivering a line. At a natural AI voiceover pace of 175 words per minute, a 30-second video only fits 85 to 90 words. Use the speech time calculator on your script before rendering. Prompts that end with three sentences of dialogue produce videos that either speed the character up unnaturally or cut off before the CTA.
Batch by category, not by day. Sit down for a session, pick one category from the library, and render five posts in a row using that category's five prompts. Consistency is easier to maintain within a category (all POV posts feel like the same show) than across a scrambled week of formats.
Never mix character descriptions inside the prompt. If your character is already saved, do not describe her face, hair, or age in the prompt. That fights the saved character and produces drift. Describe the scene; let the saved character bring the face.
The 50 prompts
Each of the 10 categories below has five prompts. Use them in sequence for a themed week (five product reviews Monday to Friday) or mix categories for variety. Every prompt is a single blockquoted line you copy, paste, and edit the bracketed pieces.
Product review (5 prompts)
The workhorse category for affiliate and brand-deal content. Each prompt gives a slightly different tone (skeptical, enthusiastic, comparison, before-and-after, honest-review) to keep the feed from feeling repetitive.
"Character [CHARACTER_NAME] wearing a casual cream sweater, standing in a bright modern kitchen, holding [PRODUCT_NAME] up to the camera, tight over-the-shoulder shot, character speaks directly to camera with a warm smile: 'I tried [PRODUCT_NAME] for 30 days and here is the honest verdict. [ONE_SPECIFIC_RESULT]. If you want the link, it is in my bio.'"
"Character [CHARACTER_NAME] wearing a soft neutral tee, seated at a wooden desk, [PRODUCT_NAME] positioned beside a coffee cup, medium shot with soft window light, character raises eyebrow skeptically then softens into a nod: 'I was ready to hate [PRODUCT_NAME]. Two weeks in, I owe it an apology. [ONE_UNEXPECTED_BENEFIT].'"
"Character [CHARACTER_NAME] wearing minimal athletic wear, standing in a well-lit bathroom mirror scene, holding [PRODUCT_NAME_A] in left hand and [PRODUCT_NAME_B] in right, tight symmetric shot, character looks left, then right, then to camera: 'Everyone asked me to compare [PRODUCT_NAME_A] vs [PRODUCT_NAME_B]. Here is what I actually noticed after using both.'"
"Character [CHARACTER_NAME] wearing a fitted white shirt, seated on a linen sofa with soft afternoon light, holding a small notebook with [PRODUCT_NAME] beside her, medium close-up, character glances at notes then delivers with quiet confidence: '[PRODUCT_NAME] is either overhyped or genuinely worth [PRICE]. I tested it against three cheaper alternatives.'"
"Character [CHARACTER_NAME] wearing a cozy oversized cardigan, sitting cross-legged on a bed with warm string lights behind, holding [PRODUCT_NAME] in her lap, cozy medium shot, character speaks in a lower, storytelling tone: 'I never thought I would recommend [PRODUCT_NAME]. Then this happened. [ONE_SENTENCE_STORY].'"
Get-ready-with-me / GRWM (5 prompts)
Runs long on retention because viewers want to see the transformation. Keep scripts to a single throughline (an event, a mood, a topic) instead of small talk.
"Character [CHARACTER_NAME] wearing a plain black tank top, seated at a vanity with soft ring light, hair partially done, applying [PRODUCT_NAME], relaxed medium close-up, character speaks casually while getting ready: 'Get ready with me for [EVENT]. While I do this I want to tell you about [TOPIC].'"
"Character [CHARACTER_NAME] wearing a robe, hair wet from the shower, standing in a bright bathroom mirror scene, reaching for [PRODUCT_NAME], mirror-reflection shot, character speaks in a fresh-morning tone: 'GRWM for a chill day at home. Today is one of those days where I have to remind myself [PIECE_OF_ADVICE].'"
"Character [CHARACTER_NAME] wearing an untied silk top, seated at a warm-lit vanity, midway through a natural makeup look, holding [PRODUCT_NAME] near her cheek, close-up on the application motion: 'Get ready with me for [EVENT]. I have been asked five times this week [COMMON_QUESTION]. Here is my actual answer.'"
"Character [CHARACTER_NAME] wearing a soft grey hoodie, seated at a minimal desk with morning light, brushing hair while [PRODUCT_NAME] sits within reach, wide medium shot, character speaks like she is on a call with a friend: 'GRWM for a slow Sunday. I want to walk you through [ROUTINE_STEP] because a lot of you asked.'"
"Character [CHARACTER_NAME] wearing black slip dress on a hanger visible behind her, seated at a glam vanity with dramatic warm light, applying [PRODUCT_NAME], slightly higher energy delivery, tighter medium shot: 'Get ready with me for a night out. Playlist is on, drink is poured, and I have a story to tell you about [TOPIC].'"
Point-of-view / POV skits (5 prompts)
Highest engagement ceiling per format. Frame the video from someone else's perspective (the barista, the friend, the stranger) so the character is reacting rather than lecturing.
"Character [CHARACTER_NAME] wearing a casual jacket, sitting across a small cafe table, camera framed as if you are seated opposite her, natural cafe ambience, character leans in slightly and speaks in confidential tone: 'POV: your friend who just tried [PRODUCT_NAME] is finally telling you what she thinks. [OPINION].'"
"Character [CHARACTER_NAME] wearing a stylish coat, standing at a shop counter with soft background blur, camera framed as if you are the shop assistant, character delivers a request with warm confidence: 'POV: a customer walks in and asks about [PRODUCT_CATEGORY]. This is exactly what I say every time. [ADVICE].'"
"Character [CHARACTER_NAME] wearing workout clothes, standing beside a gym rack with soft gym lighting, camera framed at eye level as if she just noticed you, quick smile then honest delivery: 'POV: you catch me at the gym and ask what [TOPIC] I actually recommend. Here is the real answer.'"
"Character [CHARACTER_NAME] wearing a smart-casual outfit, seated on a park bench with dappled light, camera framed as if she is talking to a friend beside her, character speaks with reflective calm: 'POV: your older sister sits you down and tells you the thing about [LIFE_TOPIC] nobody warned her about.'"
"Character [CHARACTER_NAME] wearing a comfortable knit, seated on the floor of a warm-lit apartment, camera framed as if you are seated across from her, character laughs then softens: 'POV: your best friend has to be honest with you about [SITUATION]. It is going to be a hard love kind of chat.'"
Unboxing (5 prompts)
Underrated for both engagement and sponsored posts. Keep the reveal slow. Front-load the reaction, not the product name.
"Character [CHARACTER_NAME] wearing a soft neutral outfit, seated on the floor beside a wrapped [PRODUCT_BOX], warm natural light from a nearby window, tight top-down shot on the box, character glances up with anticipation: 'I have been waiting weeks for [PRODUCT_NAME] and it is finally here. Let me show you what is inside.'"
"Character [CHARACTER_NAME] wearing a fitted white t-shirt, seated at a wooden table with [PRODUCT_BOX] centered in frame, medium shot with soft even light, character speaks calmly while pulling at the packaging: 'Unboxing [PRODUCT_NAME]. If you are on the fence, watch this before you decide.'"
"Character [CHARACTER_NAME] wearing a cozy oversized shirt, seated cross-legged on a bed with several [PRODUCT_BOXES] beside her, medium shot with warm bedroom light, character laughs and reaches for the first: 'This is the unboxing of everything I bought this month. I am going to regret at least two of these.'"
"Character [CHARACTER_NAME] wearing casual denim, standing at a kitchen island with [PRODUCT_BOX] on the counter, natural bright kitchen light, medium shot from across the island, character delivers with high curious energy: 'I paid full price for [PRODUCT_NAME]. Now I need to prove to myself it was worth it. Unboxing.'"
"Character [CHARACTER_NAME] wearing a plain sweater, seated at a small desk with a single [PRODUCT_BOX] and a coffee, close-up shot on hands opening the box, character speaks quietly like she is filming a diary entry: 'Slow unboxing of [PRODUCT_NAME]. No hype, no music, just my honest first look.'"
Storytime / life update (5 prompts)
Builds parasocial attachment fast. Keep the story to one specific incident, not a general life recap. Specifics travel.
"Character [CHARACTER_NAME] wearing a cozy hoodie, seated on a bed with dim warm lighting, medium close-up, character speaks like she is confiding: 'Storytime. Last week the strangest thing happened at [PLACE]. I have to tell someone. [OPENING_LINE_OF_STORY].'"
"Character [CHARACTER_NAME] wearing a soft cardigan, seated in a leather armchair with an ambient lamp, medium shot with warm indoor light, character sighs then smiles: 'Life update. I have been quiet for two weeks and there is a reason. [ONE_SENTENCE_REASON].'"
"Character [CHARACTER_NAME] wearing a plain tee, seated on a kitchen floor with her back against a cabinet, natural morning light, slightly lower camera angle, character speaks in a more vulnerable tone: 'Storytime about the day I finally decided to [DECISION]. If you are stuck on the same thing, this might help.'"
"Character [CHARACTER_NAME] wearing a knit sweater, seated at a cafe window seat with rain visible behind her, medium shot with soft grey light, character delivers reflective and calm: 'Storytime. Nine months ago I moved to [PLACE]. Here is what nobody tells you about the first three months.'"
"Character [CHARACTER_NAME] wearing gym clothes, seated on a bench after a workout with soft gym light, tight medium shot, character breathes then delivers with steady energy: 'Life update. Three months into [HABIT_OR_GOAL] and I want to tell you what is actually working and what is not.'"
Educational / how-to (5 prompts)
Highest save rate of any category if the value is dense. Give one specific tactic, not a full framework.
"Character [CHARACTER_NAME] wearing a smart-casual blazer, seated at a clean minimal desk with a notepad, medium shot with soft office light, character glances at notes then delivers to camera with clarity: 'If you are trying to [GOAL], here is the one mistake that costs most people six months. [SPECIFIC_MISTAKE].'"
"Character [CHARACTER_NAME] wearing a plain neutral top, standing at a whiteboard-lit wall with a small prop, medium shot, character gestures with one hand as she speaks: 'Three-step how-to for [OUTCOME]. Step one is the one everyone skips. [FIRST_STEP].'"
"Character [CHARACTER_NAME] wearing a fitted turtleneck, seated on a stool with warm even studio light, tight medium shot, character delivers with confident teacher energy: 'How to actually [TASK] without wasting a weekend. This is the exact process I use every time. [OPENING_STEP].'"
"Character [CHARACTER_NAME] wearing a soft blouse, seated at a warm cafe table with a coffee, medium shot with natural afternoon light, character speaks like she is coaching a friend: 'If you keep [PROBLEM], the fix is smaller than you think. Here is what actually worked for me. [PRACTICAL_FIX].'"
"Character [CHARACTER_NAME] wearing a casual sweater, standing in front of a bright window with a plant beside her, medium shot with soft daylight, character delivers with grounded certainty: 'The 60-second version of how I [ACHIEVEMENT]. No fluff. [FIRST_ACTION_ITEM].'"
Trending audio commentary (5 prompts)
Bends with whatever audio is trending that week. The character delivers a commentary or hot take that fits the audio's mood. Keep the delivery matched to the audio's energy.
"Character [CHARACTER_NAME] wearing a stylish jacket, standing on a city street with soft golden hour light, medium shot with slight camera movement, character delivers with wry knowing energy: '[TRENDING_AUDIO_HOOK] but the [NICHE] version. Here is what nobody says out loud. [HOT_TAKE].'"
"Character [CHARACTER_NAME] wearing casual streetwear, seated on a bench in a busy plaza with warm background blur, medium shot, character delivers with calm confidence to match a slower trending audio: '[TRENDING_AUDIO_HOOK]. If you are in [SITUATION], you already know what I am about to say. [OBSERVATION].'"
"Character [CHARACTER_NAME] wearing an oversized shirt, seated at home with cozy afternoon light, medium close-up, character speaks with playful commentary energy: 'This audio is trending but nobody is using it right for [NICHE]. Let me try. [COMMENTARY_LINE].'"
"Character [CHARACTER_NAME] wearing a smart top, standing in a lifestyle scene appropriate to [NICHE], medium shot with clean lighting, character delivers a dry one-liner then follows up: '[TRENDING_AUDIO_LINE]. The [NICHE] translation of that is [SPECIFIC_INSIGHT].'"
"Character [CHARACTER_NAME] wearing a fitted casual outfit, standing in front of a soft-colored wall, medium shot with even light, character delivers with high playful energy to match an upbeat trending audio: '[TRENDING_AUDIO_HOOK]. If you have been putting off [ACTION], take this as your sign. [SPECIFIC_STEP].'"
Reaction / react-to (5 prompts)
Works best as a second post in a series, referencing something specific your audience already saw. Keep the reaction voiced first, then explain.
"Character [CHARACTER_NAME] wearing a soft neutral top, seated at a desk with the camera framed close, warm indoor light, character raises eyebrows and delivers with genuine surprise: 'Reacting to [ARTICLE_OR_STAT]. If this is true it changes what I said last week about [TOPIC]. Here is my updated take.'"
"Character [CHARACTER_NAME] wearing a casual cardigan, seated on a sofa with soft ambient light, medium shot, character shakes head slowly then smiles: 'Reacting to [POPULAR_OPINION]. I disagree with most of it but there is one part I actually stand by. [SPECIFIC_AGREEMENT].'"
"Character [CHARACTER_NAME] wearing a fitted sweater, seated cross-legged on a bed with warm lighting, tight medium shot, character delivers with amused thoughtful energy: 'Reacting to [TREND] like everyone else. Except I have a boring take that might save you time. [BORING_BUT_TRUE_TAKE].'"
"Character [CHARACTER_NAME] wearing a plain top, standing in a bright kitchen with natural light, medium shot, character laughs briefly then delivers seriously: 'Reacting to [PIECE_OF_ADVICE] that has been going around. I tried it. Here is what actually happened. [OUTCOME].'"
"Character [CHARACTER_NAME] wearing smart-casual clothes, seated at a cafe table with a coffee, medium shot with soft window light, character delivers with measured curiosity: 'Reacting to [NEWS_OR_EVENT]. I have been thinking about this all week. My honest position is [POSITION].'"
Behind-the-scenes / a day in my life (5 prompts)
Builds long-term trust more than any other category. Show one segment of a day, not the whole thing. The specificity is the magic.
"Character [CHARACTER_NAME] wearing casual morning clothes, standing in a bright kitchen holding a coffee cup, natural morning light, medium shot with slight handheld feel, character speaks with a fresh-morning energy: 'Behind the scenes of a [DAY_OF_WEEK] in [CITY]. First thing I do is [SPECIFIC_MORNING_HABIT].'"
"Character [CHARACTER_NAME] wearing focused workwear, seated at a workspace with laptop open, warm office light, medium shot, character glances at screen then to camera: 'A day in my life if you have ever wondered what I actually do. Right now I am working on [PROJECT].'"
"Character [CHARACTER_NAME] wearing gym clothes with a small bag on shoulder, standing outside a gym in soft daylight, medium shot, character delivers with steady energy: 'A day in my life. I built this specific hour into every day for one reason. [REASON].'"
"Character [CHARACTER_NAME] wearing an apron over casual clothes, standing at a kitchen counter with ingredients laid out, warm kitchen light, medium overhead-and-front alternating shot, character speaks while prepping: 'Behind the scenes of the meal I actually eat three times a week. Ingredients are [LIST]. Here is the shortcut.'"
"Character [CHARACTER_NAME] wearing a comfortable knit, seated on a sofa with a book and warm evening lamp light, medium shot, character speaks quietly and reflectively: 'End-of-day day in my life. This is the hour I protect no matter what. Right now I am [SPECIFIC_EVENING_ACTIVITY].'"
Q&A / follower questions (5 prompts)
The compounding category. Every answered question becomes ammunition for future posts. Read the exact question aloud before answering.
"Character [CHARACTER_NAME] wearing a soft cardigan, seated at a warm-lit desk with a phone visible on the table, medium shot, character glances at phone then delivers to camera: 'Someone asked me [QUESTION]. This is a question I get almost weekly. Here is the honest answer. [ANSWER].'"
"Character [CHARACTER_NAME] wearing a plain top, seated on a stool in front of a neutral backdrop, medium shot with clean studio light, character delivers with clear teaching energy: 'Answering the top five questions from last week. Number one, [QUESTION_ONE]. My answer is [SHORT_ANSWER].'"
"Character [CHARACTER_NAME] wearing casual clothes, seated on the floor of a warm-lit living room, medium shot with a mug in hand, character delivers casually: 'Q&A round. First question is [QUESTION]. I have been thinking about how to answer this for two days. [MY_ANSWER].'"
"Character [CHARACTER_NAME] wearing a smart-casual outfit, seated at a small cafe table with a notebook, medium shot with soft afternoon light, character delivers with reflective calm: 'One of you asked [QUESTION] and I did not want to answer in a comment because it deserves a proper reply. [FULL_ANSWER].'"
"Character [CHARACTER_NAME] wearing a fitted top, seated in a neutral home office, medium close-up with soft indoor light, character delivers with confident practical energy: 'This is the question I get more than any other. [QUESTION]. Here is what I would tell my younger self. [ANSWER].'"
Common prompt mistakes
Five patterns that break consistency across a feed, even with a saved character.
1. Redescribing the character in every prompt. If the tool has a saved character, describing her face, hair colour, or age in the prompt tells the model to try to reconcile the description with the saved likeness. That fight produces drift. Describe the scene; let the saved character bring the face. This is the biggest single mistake among creators moving from generic AI video to a character-consistent workflow, and it applies whether you are using a talking-avatar renderer or a full character pipeline.
2. Changing the lighting language every post. Words like "cinematic", "golden hour", "moody", "dramatic" pull renders in wildly different directions. Pick one lighting family (warm natural, soft even studio, cool ambient) and use the same language across your library. Save the dramatic-lighting swings for special posts, not the daily feed.
3. Cramming three shot types into one prompt. Prompts that ask for "wide establishing then medium then close-up" produce muddled renders on most current tools. Pick one shot type per prompt. If you need cut coverage, render three separate prompts and edit them together in a script-to-video pipeline.
4. Writing dialogue that is too long. Every prompt in this library ends with the character delivering a line. A 30-second clip only fits about 85 words. Prompts with 200-word monologues either get rushed unnaturally or cut off mid-sentence. Trim the dialogue to the length the video will actually be.
5. Using different tools within the same account. Consistency does not survive tool-hopping. A character rendered in Tool A on Monday and Tool B on Wednesday looks like two different people, because the underlying models learn features slightly differently. Pick one tool, ship the first 30 posts on it, then evaluate before switching.
A prompt library is only the second half of consistency. The first half is the saved character and the locked voice. The prompt library layer sits on top of those two. Without them, the best prompt still produces a lookalike.
Frequently asked questions
What is an AI influencer prompt library?
An AI influencer prompt library is a reusable set of scene templates that keep your character, voice, and framing consistent across dozens of posts. Each prompt has fixed skeleton language and bracketed placeholders for the specific script, product, or setting. You paste the template, swap the placeholders, and render.
Why should I use the same prompt structure across posts?
Because followers recognise creators through repetition. The same lighting language, shot type, and delivery style across 30 posts trains the algorithm and the audience to identify your creator. Rewriting from scratch each time fights that recognition.
Can these prompts be used with any AI video tool?
The bracketed prompt structure works with most character-consistent AI video tools in 2026, including MakeAIVideo's character pipeline, Higgsfield, and The Influencer AI. Tools without a saved-character feature will still drift no matter how well you prompt.
How long should each prompt's dialogue be?
About 85 to 90 words for a 30-second clip, or 130 words for a 45-second clip. At the natural AI voiceover pace of 175 words per minute, longer dialogue gets rushed or truncated. Trim ruthlessly.
Do I need to describe the character in every prompt?
No. If your tool has a saved character, redescribing her face in the prompt actively hurts consistency by making the model reconcile the description with the saved likeness. Describe only the scene, outfit, and lighting.
How many prompts should I ship before evaluating an account?
At least 30 posts on a single platform with consistent cadence. Fewer posts and the algorithm has not built a recommendation profile, so the results are noise. Some hook patterns emerge earlier, but account-level performance takes 30 posts to read.
Are trending-audio prompts safe from copyright issues?
The character is generated so the visual side is safe, but the trending audio itself follows the same rules as any other creator's use. Use the platform's licensed audio library (Instagram's, TikTok's, YouTube Shorts') so licensing is handled by the platform. Adding your own commercially licensed audio is the other safe path.
How is a prompt library different from a script generator?
A prompt library gives you the visual scene template (character, lighting, shot, delivery) with a script slot. A script generator gives you the script but not the scene. Use both: the video script generator writes the words; this library places them into a repeatable visual container.
Should I disclose that these posts are AI-generated?
Yes. Meta's community standards on synthetic media, TikTok's integrity and authenticity guidelines, and YouTube's Partner Program policies all require disclosure of AI-generated content in 2026, and sponsored posts additionally require FTC-compliant disclosure of the paid partnership. Tick the AI-generated toggle at upload; it does not affect reach.
What voice should I use with these prompts?
One voice for the entire library. Pick a voice that matches the character's apparent age and energy, save it, and use it on every render. A benchmark like ElevenLabs has voice families in this consistency band; most current AI-influencer tools ship a curated voice library that works similarly. The rule is: one character, one voice, forever.
Where to start
Pick one category from the library and ship five posts using its prompts before you branch out. The point of a prompt library is not variety on day one; it is compounding recognition over weeks. Five product-review posts in a row teach the algorithm and your followers to identify the character faster than five scattered posts across five categories.
If you have not picked a tool yet, our character-driven pipeline ships the three pieces this library depends on: saved character, locked voice, and platform-native rendering. Plans start at $29/month with a 7-day free trial, $0 due today, cancel anytime in the trial window for no charge. For the full step-by-step from zero to a posting-ready creator, the complete AI influencer guide is the walkthrough that pairs with this library.
Save the character, pick a category, ship the first five. That is the whole workflow.