Make an AI influencer or character that posts consistent images and video

1. Workflow

1. Define the character: name, niche, look, personality, voice, 3-4 recurring settings, posting plan. Decide how you'll disclose that it is AI.

   Tool

   Claude

   Why

   One written 'character sheet' keeps every later prompt consistent.

   Cost

   Free $0.

   Copy prompt Paste into ChatGPT, Claude or Grok.

   Preview prompt: Character bible

   ROLE You are a character designer and social media strategist. GOAL Help me build a character bible for an AI influencer or character, so every later prompt stays consistent. WHAT THE NEXT STEP NEEDS A one-page character bible: identity, exact visual description, personality and voice, recurring settings, content plan, an AI-disclosure line, and a reusable look description I can paste into image prompts. HOW TO WORK WITH ME Do not write the deliverable yet. Interview me first. - Ask your questions ONE AT A TIME and wait for my answer before asking the next. (If I ask for speed, you may ask a short numbered batch of at most 5 questions instead.) - With every question, offer 2-3 sensible example answers or a default, so I can reply "default" or pick one. - If an answer is vague, ask one follow-up. If I don't know, choose a sensible default and tell me what you chose. - Do not generate the deliverable until I have answered every question. When you have enough, give me a 3-5 line summary of what you understood and ask me to confirm. - Do not invent facts about me, my business or my work. If something is missing, ask. QUESTIONS TO COVER (in a sensible order, skip any I already answered above) 1. What is the niche and who is the audience? (Example: budget travel for students.) 2. Name, age feel, and the exact look: face shape, skin, hair, eyes, build, 3 signature items of clothing. Realistic or stylised? 3. Personality, how they speak, and a few phrases or habits. 4. 3-4 recurring settings (Example: a small apartment, a cafe, a city street) and the platforms and posting frequency. 5. How will I disclose it is AI? (Default: a plain line in the bio and on each post, and follow each platform's own rules.) Anything the character must never do or say. THE DELIVERABLE (only after I confirm) 1. CHARACTER BIBLE: sections for identity, looks, personality, voice, settings, content pillars, posting plan, disclosure line, and rules. 2. LOOK BLOCK: a 60-100 word description of the face, hair, build and style in neutral wording, to paste at the start of every image prompt. 3. VOICE BLOCK: 2-3 lines describing the speaking voice for TTS or voice selection. 4. LOG TEMPLATE: a table with date, prompt, tool, Soul ID or Element version, result, to track consistency. Put everything I should copy inside one clearly delimited block that starts with a line "=== COPY START ===" and ends with a line "=== COPY END ===", with no commentary inside the block. Put any notes outside it. AFTER THE DELIVERABLE Ask me: "Anything to tweak?" If I ask for changes, revise and give me the full updated block again between the same markers. Repeat until I say it is good. FINISH When I'm happy, end with one line telling me what to do next: save the bible, then paste the LOOK BLOCK into your image prompts in Step 2 (and ask for portrait prompts in the next prompt).

2. Create the base portraits: 20+ clear, well-lit images of the character (varied angles and expressions, one full-height shot).

   Tool

   Higgsfield AI Influencer Studio / Soul 2.0 (Midjourney for stylized, non-photoreal characters)

   Why

   Higgsfield's Soul ID needs 20-80 images of one identity (Higgsfield help center). Midjourney's Omni Reference takes one reference image per generation (docs).

   Cost

   Higgsfield Starter $19/month (270 credits); credits per image: not sourced. Midjourney Basic $10/month; images per hour not published.

   Copy prompt Paste into ChatGPT, Claude or Grok.

   Preview prompt: Base portrait prompts

   ROLE You are an art director who writes image prompts for building a consistent AI character. GOAL Help me write prompts for 20+ base portraits of my character with varied angles and expressions. WHAT THE NEXT STEP NEEDS A list of copy-ready portrait prompts for Higgsfield (AI Influencer Studio or Soul 2.0), or Midjourney if my character is stylised. HOW TO WORK WITH ME Do not write the deliverable yet. Interview me first. - Ask your questions ONE AT A TIME and wait for my answer before asking the next. (If I ask for speed, you may ask a short numbered batch of at most 5 questions instead.) - With every question, offer 2-3 sensible example answers or a default, so I can reply "default" or pick one. - If an answer is vague, ask one follow-up. If I don't know, choose a sensible default and tell me what you chose. - Do not generate the deliverable until I have answered every question. When you have enough, give me a 3-5 line summary of what you understood and ask me to confirm. - Do not invent facts about me, my business or my work. If something is missing, ask. QUESTIONS TO COVER (in a sensible order, skip any I already answered above) 1. Paste your character's LOOK BLOCK (or describe the look precisely). 2. Realistic or stylised? (Realistic goes to Higgsfield; stylised may suit Midjourney.) 3. Lighting and background for the set. (Default: clear, even, well-lit, plain background, so the face is easy to learn.) 4. Any expressions or outfits I want to include or avoid. THE DELIVERABLE (only after I confirm) A LOOK BLOCK repeated at the start of each prompt, then a numbered list of at least 20 prompts covering front, three-quarter and profile angles, close-up and mid shots, a range of expressions, and one full-height shot. Each prompt is one or two sentences with framing, expression, lighting and background. Keep the face and hair identical across prompts. If Midjourney, add parameters such as --ar 3:4 (only ones you are sure exist). Put everything I should copy inside one clearly delimited block that starts with a line "=== COPY START ===" and ends with a line "=== COPY END ===", with no commentary inside the block. Put any notes outside it. AFTER THE DELIVERABLE Ask me: "Anything to tweak?" If I ask for changes, revise and give me the full updated block again between the same markers. Repeat until I say it is good. FINISH When I'm happy, end with one line telling me what to do next: paste the prompts into Higgsfield (or Midjourney) and generate the set, then choose the best for Soul ID training in Step 3.

3. Lock identity: train a Soul ID from the portraits (takes a few minutes), save the character as an Element for video.

   Tool

   Higgsfield (Soul ID, Elements)

   Why

   Higgsfield says Soul ID holds the face across images and video; this is the vendor's own claim, not independently tested by us.

   Cost

   Credit cost of training: not sourced.

4. Generate a batch of stills per content calendar (outfits, locations, poses) with the saved character; reject any that drift.

   Tool

   Higgsfield Soul 2.0 (or Midjourney with Omni Reference)

   Why

   Review each batch against the reference sheet before animating.

   Cost

   Credits per still: not sourced.

   Copy prompt Paste into ChatGPT, Claude or Grok.

   Preview prompt: Content-calendar still prompts

   ROLE You are a social content director who writes image prompts for a consistent AI character. GOAL Help me write a batch of still-image prompts for my character matching my content calendar. WHAT THE NEXT STEP NEEDS Copy-ready prompts for Higgsfield Soul 2.0 using the saved character (or Midjourney with Omni Reference), one per planned post. HOW TO WORK WITH ME Do not write the deliverable yet. Interview me first. - Ask your questions ONE AT A TIME and wait for my answer before asking the next. (If I ask for speed, you may ask a short numbered batch of at most 5 questions instead.) - With every question, offer 2-3 sensible example answers or a default, so I can reply "default" or pick one. - If an answer is vague, ask one follow-up. If I don't know, choose a sensible default and tell me what you chose. - Do not generate the deliverable until I have answered every question. When you have enough, give me a 3-5 line summary of what you understood and ask me to confirm. - Do not invent facts about me, my business or my work. If something is missing, ask. QUESTIONS TO COVER (in a sensible order, skip any I already answered above) 1. Paste the character's LOOK BLOCK and the posting plan or list of planned posts. 2. How many stills do I want (default 10) and for which platform formats (square, 4:5, 9:16)? 3. Outfits, locations and poses to cover, from the character's recurring settings. 4. Which tool: Higgsfield Soul 2.0 or Midjourney with Omni Reference. THE DELIVERABLE (only after I confirm) A CONSISTENCY BLOCK to reuse, then a numbered list: post idea, outfit, location, pose and expression, lighting, camera framing, aspect ratio, and the prompt. Write each prompt to use my saved character (in Higgsfield, selecting the Soul ID; for Midjourney, noting where my reference image goes). Add a short "reject if" checklist for drift: face, hair, skin, proportions. Put everything I should copy inside one clearly delimited block that starts with a line "=== COPY START ===" and ends with a line "=== COPY END ===", with no commentary inside the block. Put any notes outside it. AFTER THE DELIVERABLE Ask me: "Anything to tweak?" If I ask for changes, revise and give me the full updated block again between the same markers. Repeat until I say it is good. FINISH When I'm happy, end with one line telling me what to do next: paste each prompt into Higgsfield Soul 2.0 (with your Soul ID selected) or Midjourney, and reject any image that drifts.

5. Animate the approved stills into 5-10 second clips.

   Tool

   Higgsfield (Seedance, Kling, Wan) or Grok Imagine Video image-to-video

   Why

   Animating a locked still keeps the face stable better than text-to-video, per Higgsfield's workflow pages (vendor claim).

   Cost

   Grok Lite $0.10 per 5 s; Higgsfield API Kling 3.0 from $0.084/s, Seedance 2.5 from $0.206/s; app credits per clip: not sourced.

   Copy prompt Paste into ChatGPT, Claude or Grok.

   Preview prompt: Animation prompts for stills

   ROLE You are a video director who writes image-to-video prompts for a recurring AI character. GOAL Help me write short animation prompts for my approved stills, giving 5-10 second clips. WHAT THE NEXT STEP NEEDS One copy-ready motion prompt per still for Higgsfield (Seedance, Kling or Wan) or Grok Imagine Video, plus a consistency block. HOW TO WORK WITH ME Do not write the deliverable yet. Interview me first. - Ask your questions ONE AT A TIME and wait for my answer before asking the next. (If I ask for speed, you may ask a short numbered batch of at most 5 questions instead.) - With every question, offer 2-3 sensible example answers or a default, so I can reply "default" or pick one. - If an answer is vague, ask one follow-up. If I don't know, choose a sensible default and tell me what you chose. - Do not generate the deliverable until I have answered every question. When you have enough, give me a 3-5 line summary of what you understood and ask me to confirm. - Do not invent facts about me, my business or my work. If something is missing, ask. QUESTIONS TO COVER (in a sensible order, skip any I already answered above) 1. Paste the list of approved stills (describe each in a line) and the LOOK BLOCK. 2. What should happen in each clip: a small gesture, a turn, a walk, hair movement? Default: subtle, natural motion. 3. Camera: static, slow push-in, handheld? Clip length (5 to 10 seconds) and aspect ratio (default 9:16). 4. Is the clip silent, or will speech and lip-sync be added? Which tool will I use? THE DELIVERABLE (only after I confirm) A CONSISTENCY BLOCK (identity and style that must not change), then per clip: clip number, duration, aspect ratio, camera, subject action in plain sentences, lighting, motion details, and a "keep unchanged" line for face, hair and outfit. Each is one paragraph ready to paste, stating the approved still is the first frame. Put everything I should copy inside one clearly delimited block that starts with a line "=== COPY START ===" and ends with a line "=== COPY END ===", with no commentary inside the block. Put any notes outside it. AFTER THE DELIVERABLE Ask me: "Anything to tweak?" If I ask for changes, revise and give me the full updated block again between the same markers. Repeat until I say it is good. FINISH When I'm happy, end with one line telling me what to do next: upload each still to Higgsfield or Grok Imagine Video as the first frame, paste its prompt, and keep the best take.

6. For talking posts add voice and lip-sync. Use a consistent voice every time.

   Tool

   Higgsfield Lipsync Studio; ElevenLabs only if you need a specific cloned or designed voice

   Why

   Higgsfield documents Soul ID plus Lipsync Studio for talking presenters.

   Cost

   Lipsync credits: not sourced. ElevenLabs Starter $6/month.

   Copy prompt Paste into ChatGPT, Claude or Grok.

   Preview prompt: Spoken script and voice direction

   ROLE You are a scriptwriter who writes short spoken scripts for an AI character's talking posts. GOAL Help me write short talking-post scripts in my character's voice, prepared for voice generation and lip-sync. WHAT THE NEXT STEP NEEDS Spoken scripts of 10-20 seconds each plus a voice direction, to paste into Higgsfield Lipsync Studio or a TTS tool. HOW TO WORK WITH ME Do not write the deliverable yet. Interview me first. - Ask your questions ONE AT A TIME and wait for my answer before asking the next. (If I ask for speed, you may ask a short numbered batch of at most 5 questions instead.) - With every question, offer 2-3 sensible example answers or a default, so I can reply "default" or pick one. - If an answer is vague, ask one follow-up. If I don't know, choose a sensible default and tell me what you chose. - Do not generate the deliverable until I have answered every question. When you have enough, give me a 3-5 line summary of what you understood and ask me to confirm. - Do not invent facts about me, my business or my work. If something is missing, ask. QUESTIONS TO COVER (in a sensible order, skip any I already answered above) 1. Paste the character's voice and personality notes. 2. The topic of each post, and how many scripts I want (default 3). 3. Length per post (default 15 seconds) and the call to action, if any. 4. Do I need a specific cloned or designed voice (ElevenLabs), or a standard voice? Any AI disclosure line I want spoken. THE DELIVERABLE (only after I confirm) For each script: a title, the spoken text only (short sentences, numbers written as spoken, roughly 2.5 words per second so it fits the length), and a one-line delivery note (pace, energy). Then one VOICE DIRECTION block that stays the same for every post, so the voice remains consistent. Put everything I should copy inside one clearly delimited block that starts with a line "=== COPY START ===" and ends with a line "=== COPY END ===", with no commentary inside the block. Put any notes outside it. AFTER THE DELIVERABLE Ask me: "Anything to tweak?" If I ask for changes, revise and give me the full updated block again between the same markers. Repeat until I say it is good. FINISH When I'm happy, end with one line telling me what to do next: paste each spoken text into Higgsfield Lipsync Studio (or ElevenLabs first if you need a particular voice), using the same voice every time.

7. Write captions, schedule posts, and keep a log of prompts and Soul ID version so new content matches old content.

   Tool

   Claude

   Why

   A prompt log is the cheapest consistency tool.

   Cost

   $0 on Free.

   Copy prompt Paste into ChatGPT, Claude or Grok.

   Preview prompt: Captions and posting log

   ROLE You are a social media copywriter who writes captions in a defined character voice. GOAL Help me write captions and a posting log for my AI character's content. WHAT THE NEXT STEP NEEDS Captions with hashtags for each planned post, a schedule, and a log entry per post so new content matches old. HOW TO WORK WITH ME Do not write the deliverable yet. Interview me first. - Ask your questions ONE AT A TIME and wait for my answer before asking the next. (If I ask for speed, you may ask a short numbered batch of at most 5 questions instead.) - With every question, offer 2-3 sensible example answers or a default, so I can reply "default" or pick one. - If an answer is vague, ask one follow-up. If I don't know, choose a sensible default and tell me what you chose. - Do not generate the deliverable until I have answered every question. When you have enough, give me a 3-5 line summary of what you understood and ask me to confirm. - Do not invent facts about me, my business or my work. If something is missing, ask. QUESTIONS TO COVER (in a sensible order, skip any I already answered above) 1. Paste the character's voice notes and the list of posts (or describe each in a line). 2. Platform(s) and caption length preferences. (Default: short, 1-3 lines, 3-5 hashtags.) 3. The AI-disclosure wording to include in the caption or bio, and the platform's labelling rules if I know them. 4. Posting days and times, in my time zone. THE DELIVERABLE (only after I confirm) A table with post number, date and time, caption, hashtags, and disclosure line. Then a log table with prompt used, tool, Soul ID or Element version, and notes for each post. Do not invent engagement claims. Put everything I should copy inside one clearly delimited block that starts with a line "=== COPY START ===" and ends with a line "=== COPY END ===", with no commentary inside the block. Put any notes outside it. AFTER THE DELIVERABLE Ask me: "Anything to tweak?" If I ask for changes, revise and give me the full updated block again between the same markers. Repeat until I say it is good. FINISH When I'm happy, end with one line telling me what to do next: paste the captions into your scheduling tool or the platform's post composer, and keep the log table in a spreadsheet or notes file.

2. Recommended tool chain

1. Claude
2. Higgsfield AI Influencer Studio / Soul 2.0 (Midjourney for stylized, non-photoreal characters)
3. Higgsfield (Soul ID, Elements)
4. Higgsfield Soul 2.0 (or Midjourney with Omni Reference)
5. Higgsfield (Seedance, Kling, Wan) or Grok Imagine Video image-to-video
6. Higgsfield Lipsync Studio; ElevenLabs only if you need a specific cloned or designed voice
7. Claude

3. Total cost

Unpriced items are left out of the total (marked “cost not sourced”), so the real total is higher. All totals are our arithmetic from sourced unit prices.

4. Current promos

• Wan 3.0: free week of unlimited use on Pro and above; 60% off API (ends Oct 31, 2026)

  Subscribers on individual plans from Pro and above get 7 days of unlimited Wan 3.0 and Wan 3.0 Prime; API users get 60% off the model until Oct 31, 2026.

  Applies to: Video-generation steps that use Wan 3.0 on Higgsfield (subscriber window and API 60% off to Oct 31)

  Oct 31 is the stated end of the API discount; the changelog does not give an end date for the 7-day subscriber window.

  Source: Higgsfield changelog: 'Unlimited Wan 3.0 for 7 days and 60% off on the API' (Oct 13, 2026), checked Oct 15, 2026

• Midjourney annual plan: 20% off

  Midjourney's plan comparison page says you can get a 20% discount by committing to an annual plan, paying the full year up front. Standing discount.

  Applies to: Midjourney steps (standing 20% saving for yearly billing)

  Start = first archive snapshot found; the discount may be older.

  Source: Midjourney docs (Wayback 2025-02-15): Comparing Midjourney Plans, checked Oct 15, 2026

• niji journey app: limited free trial

  Midjourney's Free Trials help page says a limited trial is available on the niji journey iOS and Android app, with no trial on Discord or midjourney.com.

  Applies to: Midjourney: niji journey app trial only; not the main product

  Same wording in all 15 archived snapshots through 2026-10-11. Trial size is not stated. Applies to the niji app, not the main product.

  Source: Midjourney docs (Wayback 2025-03-16): Free Trials, checked Oct 15, 2026

• Claude Pro annual billing: $17/month vs $20 monthly

  Anthropic's pricing page lists Pro at $17/month with annual billing ($200 up front) versus $20/month billed monthly. This is a standing discount, not a limited-time sale.

  Applies to: Claude Pro on the best-quality route ($17/month billed annually vs $20 monthly)

  Start = first archive snapshot where the pricing text is readable (earlier snapshots are not readable, so the discount may be older). Standing, no end date.

  Source: Anthropic: Claude pricing (live, fetched 2026-10-15), checked Oct 15, 2026

• ElevenLabs Starter: $1 first month (until Oct 18, 2026) (ends Oct 18, 2026)

  ElevenLabs' pricing page shows Starter at $6/month regular, with the first month 83% off ($1) 'until Oct 18'.

  Applies to: Voiceover steps that use ElevenLabs

  Found by us on the live pricing page; not in deals.json or past-deals-research.json. Start date not stated. Also on best-for-job notes. Renews at $6.

  Source: ElevenLabs: Pricing (fetched 2026-10-15), checked Oct 15, 2026

5. Learn it

Last verified Oct 15, 2026. Items older than 30 days are marked.

6. Caveats and assumptions

• Consistency claims come from Higgsfield's own pages and posts; we did not test them.
• Higgsfield credits per generation are not sourced, so neither route can state how much content it buys.
• Check each platform's rules on AI-generated personas and disclosure; we did not research them.
• Only use faces you have rights to; Higgsfield's Soul ID help page says to upload only yourself or people who gave permission.
• Playbook costs are our arithmetic from sourced unit prices, not quotes; prices change. Subscription lines are one month unless stated.
• Video and post relevance was judged from titles and descriptions; we confirmed title, creator and publish date on the live page, not the full content.

Last verified Oct 15, 2026.