Mpromp2 — the AI Prompt GeneratorPrompt Library / Image & Video / Character Keyframe Consistency Workflow
Character Keyframe Consistency Workflow Prompt for Midjourney, DALL-E, FLUX & Ideogram
This is a free, ready-made prompt for midjourney, dall-e, flux & ideogram that lock one character's face, hair and wardrobe into a reusable keyframe sheet, so every later image and video generation can be matched against it. Copy it, replace the bracketed parts, and paste it into your AI — no login and no cost.
The Midjourney prompt
Push copies this prompt with a reference line naming your image, and downloads the image file — attach it in the model to complete the pair. This page stages the pair for you; it cannot upload into the model itself (sites cannot reach into each other), so the final attach happens there. The image is downscaled to at most 1024px and stored only in this browser, paired with this prompt. If browser storage is full, it is kept for this visit only.
The DALL-E prompt
Push copies this prompt with a reference line naming your image, and downloads the image file — attach it in the model to complete the pair. This page stages the pair for you; it cannot upload into the model itself (sites cannot reach into each other), so the final attach happens there. The image is downscaled to at most 1024px and stored only in this browser, paired with this prompt. If browser storage is full, it is kept for this visit only.
The FLUX prompt
Push copies this prompt with a reference line naming your image, and downloads the image file — attach it in the model to complete the pair. This page stages the pair for you; it cannot upload into the model itself (sites cannot reach into each other), so the final attach happens there. The image is downscaled to at most 1024px and stored only in this browser, paired with this prompt. If browser storage is full, it is kept for this visit only.
The Ideogram prompt
Push copies this prompt with a reference line naming your image, and downloads the image file — attach it in the model to complete the pair. This page stages the pair for you; it cannot upload into the model itself (sites cannot reach into each other), so the final attach happens there. The image is downscaled to at most 1024px and stored only in this browser, paired with this prompt. If browser storage is full, it is kept for this visit only.
The Stable Diffusion prompt
Push copies this prompt with a reference line naming your image, and downloads the image file — attach it in the model to complete the pair. This page stages the pair for you; it cannot upload into the model itself (sites cannot reach into each other), so the final attach happens there. The image is downscaled to at most 1024px and stored only in this browser, paired with this prompt. If browser storage is full, it is kept for this visit only.
The Nano Banana (Gemini) prompt
Push copies this prompt with a reference line naming your image, and downloads the image file — attach it in the model to complete the pair. This page stages the pair for you; it cannot upload into the model itself (sites cannot reach into each other), so the final attach happens there. The image is downscaled to at most 1024px and stored only in this browser, paired with this prompt. If browser storage is full, it is kept for this visit only.
The Adobe Firefly prompt
Push copies this prompt with a reference line naming your image, and downloads the image file — attach it in the model to complete the pair. This page stages the pair for you; it cannot upload into the model itself (sites cannot reach into each other), so the final attach happens there. The image is downscaled to at most 1024px and stored only in this browser, paired with this prompt. If browser storage is full, it is kept for this visit only.
The Leonardo prompt
Push copies this prompt with a reference line naming your image, and downloads the image file — attach it in the model to complete the pair. This page stages the pair for you; it cannot upload into the model itself (sites cannot reach into each other), so the final attach happens there. The image is downscaled to at most 1024px and stored only in this browser, paired with this prompt. If browser storage is full, it is kept for this visit only.
The Recraft prompt
Push copies this prompt with a reference line naming your image, and downloads the image file — attach it in the model to complete the pair. This page stages the pair for you; it cannot upload into the model itself (sites cannot reach into each other), so the final attach happens there. The image is downscaled to at most 1024px and stored only in this browser, paired with this prompt. If browser storage is full, it is kept for this visit only.
The Krea prompt
Push copies this prompt with a reference line naming your image, and downloads the image file — attach it in the model to complete the pair. This page stages the pair for you; it cannot upload into the model itself (sites cannot reach into each other), so the final attach happens there. The image is downscaled to at most 1024px and stored only in this browser, paired with this prompt. If browser storage is full, it is kept for this visit only.
The Grok Imagine prompt
Push copies this prompt with a reference line naming your image, and downloads the image file — attach it in the model to complete the pair. This page stages the pair for you; it cannot upload into the model itself (sites cannot reach into each other), so the final attach happens there. The image is downscaled to at most 1024px and stored only in this browser, paired with this prompt. If browser storage is full, it is kept for this visit only.
The Seedream prompt
Push copies this prompt with a reference line naming your image, and downloads the image file — attach it in the model to complete the pair. This page stages the pair for you; it cannot upload into the model itself (sites cannot reach into each other), so the final attach happens there. The image is downscaled to at most 1024px and stored only in this browser, paired with this prompt. If browser storage is full, it is kept for this visit only.
The Qwen-Image prompt
Push copies this prompt with a reference line naming your image, and downloads the image file — attach it in the model to complete the pair. This page stages the pair for you; it cannot upload into the model itself (sites cannot reach into each other), so the final attach happens there. The image is downscaled to at most 1024px and stored only in this browser, paired with this prompt. If browser storage is full, it is kept for this visit only.
The HunyuanImage prompt
Push copies this prompt with a reference line naming your image, and downloads the image file — attach it in the model to complete the pair. This page stages the pair for you; it cannot upload into the model itself (sites cannot reach into each other), so the final attach happens there. The image is downscaled to at most 1024px and stored only in this browser, paired with this prompt. If browser storage is full, it is kept for this visit only.
The MAI-Image prompt
Push copies this prompt with a reference line naming your image, and downloads the image file — attach it in the model to complete the pair. This page stages the pair for you; it cannot upload into the model itself (sites cannot reach into each other), so the final attach happens there. The image is downscaled to at most 1024px and stored only in this browser, paired with this prompt. If browser storage is full, it is kept for this visit only.
The Z-Image prompt
Push copies this prompt with a reference line naming your image, and downloads the image file — attach it in the model to complete the pair. This page stages the pair for you; it cannot upload into the model itself (sites cannot reach into each other), so the final attach happens there. The image is downscaled to at most 1024px and stored only in this browser, paired with this prompt. If browser storage is full, it is kept for this visit only.
The Kandinsky prompt
Push copies this prompt with a reference line naming your image, and downloads the image file — attach it in the model to complete the pair. This page stages the pair for you; it cannot upload into the model itself (sites cannot reach into each other), so the final attach happens there. The image is downscaled to at most 1024px and stored only in this browser, paired with this prompt. If browser storage is full, it is kept for this visit only.
The deliverable — and how to check it
Deliverable: one reference sheet image containing exactly 4 views of the same character — front portrait, three-quarter portrait, full-body front, full-body profile — with identical face, hairstyle and wardrobe in every view. Acceptance check: compare the four faces side by side; face shape, hairline, and outfit details must match in all four. If any view drifts, regenerate using the strongest view as the reference before using the sheet — a keyframe with drift in it poisons every generation matched to it.
Part of the Biography Film Production workflow
Step 1 of 3 — lock the subject
This prompt is step 1 of the Biography Film Production workflow — the output package is a biography or documentary film about a real person. Everything character-based in the film depends on this keyframe, so it comes first: no consistent likeness exists until this sheet is made and approved. Next in the chain: the Portrait Consistency Set (step 2) builds the subject's expression range from this keyframe, and the Talking-Head / Avatar Setup Shot (step 3) puts the subject on camera. The keyframe made here is also the reference image you attach to those later prompts.
This prompt for your model
How to use this prompt
Replace the bracketed placeholders — [CHARACTER DESCRIPTION], [WARDROBE], [SETTING OR BACKDROP] — with your own details, then paste the whole prompt as your first message. The bracketed parts are the only editing needed; the instructions around them are what make the answer come back structured instead of vague. If the first result is close but not right, ask the AI to adjust one thing at a time rather than rewriting the prompt.
Want a different version of this prompt?
Open the prompt gateway and type your goal in one line — it builds a fresh prompt around your exact wording, formatted for whichever model you run, with your choice of tone and output format.
Related prompts
Mpromp2 — the AI Prompt Generator. Every prompt in this library is free to use, including commercially.