🎬 Reel Maker — Team Guide

Turn a short written story + your photos and clips into a finished, captioned vertical video — no editing skills needed. This guide explains every button and field in plain language, with the exact words you can type.

What's inside

1 · What it does (in one minute)

You give it a story written as short lines and a folder of photos/videos. For each line it decides what should be on screen, how the camera moves, and the mood — then builds one 9:16 vertical video with captions and music, ready for Reels / Shorts.

Think of it as an automated video editor with a film director's brain. Each line of your story becomes one shot. A 10-line story becomes a 10-shot reel.

2 · The 8 steps, start to finish

3 · Step 1 — Subject (optional)

4 · Step 2 — Write the story

The story is plain text. Start each shot with the word Frame and a number, then write the line that should appear on screen. That's it.

Copy-paste template

Reels

Frame 1
From a farmer to a model… the story of a desi girl with big dreams.

Frame 2
At 19, I got married into a traditional Rajasthani home.

Frame 3
A kisan ki beti — I rode tractors even after marriage.

Caption:
Write your full Instagram caption here. (This does NOT appear in the video.)

Adding your own photos & videos

In story mode, start with either I already have a frame script for full control, or I have a story to let the AI create an editable frame draft from raw notes/transcript. The AI does not render directly from the raw story; it fills the script/frame cards so you can review first.

Click 📁 Browse folder… (next to the start panel) and pick the folder on your computer that holds the photos/videos for this story. The tool uploads them and matches the right one to each shot automatically.

• Put all the media for one story in a single folder first, then browse to it.
• Supported: JPG, PNG, HEIC, WEBP images and MP4, MOV, M4V, WEBM videos.
• The folder browser works on any device (phone/tablet/desktop) — no file size limit.
• You can still change the picture on any individual shot afterwards (see Picture source below).
• No folder? Leave it empty and the tool will create every visual with AI.

5 · The Frame Card — every field explained

After Parse Frames →, each line becomes a card. Top to bottom you choose: the picture, then optionally a mood note, a photo edit, a camera move, a lip-sync toggle, and the length.

5a · Picture source — where the shot's image comes from

5b · Director Note — the most important box

This tells the AI director the feeling and look of the shot. Write it like you're briefing a photographer: what's in frame + how they feel + the light.

5c · Camera Motion — pick from the dropdown

Camera is a dropdown, not a text box — no typing. It's pre-set to the AI's best pick for that beat (you'll see a ✨ auto note). To change it, just choose another move from the list below. Leave it on ✨ Auto to let the director decide at render. After 👁 Preview Stills, the ✨ Suggest from image button re-picks the best move by looking at the actual frame. Here's what each move does:

5d · Image Edit — change a photo with words

The ✏️ Image edit box changes the picture before it animates. Type a plain instruction.

Big mood changes ("add storm", "warmer light") work great. Precise moves ("shift her to the left") are less reliable — that's expected.

5e · Lip Sync — make the subject speak

Tick 🎙 Lip Sync and that shot becomes a talking face that says the line aloud in a real voice. A voice menu appears — pick one or leave the default.

• What gets spoken = the line you wrote for that frame.
• Length is automatic — the shot lasts exactly as long as the spoken line (the duration box greys out).
• Music ducks under the voice automatically.
• Safe: if a voice/lip-sync service is unavailable, the shot quietly becomes a normal animated photo. Nothing breaks.

5f · Voice-Over — narration over the full video

Check ☑ Voice-Over Track at the top to make the entire video narrated. Each caption gets read aloud in a chosen voice, creating a full voice-over narration layered under the music.

• Every caption is spoken — each line reads exactly as long as the shot lasts, no gaps.
• Voice menu — same voices as Lip Sync; pick one or use the default.
• Plays over music — narration sits on top; music stays at 25% volume behind the voice.
• No lip-sync — the voice is heard but faces don't move. Use this for cinematic storytelling, not dialogue.
• Safe: if narration fails, the video still renders with just music.

5g · Duration & Video Start

5h · Picture & Video "model" pickers

"Model" = which AI engine makes the picture or the motion. Leave both on Auto and the tool picks the best one for each shot, cheap while testing and premium for the final. Only override if you want a specific look.

Picture models

Video (motion) models

6 · Style, Quality, Mood & Captions

Quality tier — "test cheap, finish expensive"

Mood / colour palette

Orientation, transition & captions

7 · Music

Good auto-music prompts: Emotional Bollywood instrumental, struggle to triumph, sitar and tabla, no lyrics · Melancholic Rajasthani folk, raw acoustic

Combine music + voice-over: Enable Voice-Over Track (above) to layer narration on top of your music. Music plays at 25% volume, so the voice is clear and the music sits underneath.

8 · The 💰 cost panel — read it before Generate

After you parse, a cost estimate appears and updates live as you change settings. Check it before pressing Generate. Want it cheaper? Use Dev tier, prefer real photos/videos, and keep lip-sync to a few shots.

8b · Iterate without re-rendering — redo, approve, live clips

You don't have to render the whole video and hope. Treat the tool as a draft machine you steer: preview the stills, fix the frames you don't like, approve the rest, and only pay to animate what you've approved. Three controls make this fast.

🔄 Redo still — fix one image, not the whole video

After Preview Stills, each frame card shows a 🔄 Redo still button. Change that frame's note, photo, image edit, or camera — then click 🔄. Only that one image regenerates (a few seconds), and it swaps into the card. Nothing else is touched, and you don't pay to re-animate anything.

• Use it to dial in a face, a mood, or a symbolic shot until that single frame is right.
• Unchanged frames are never re-touched, so it's cheap and fast.
• When the still looks right, it's the exact image that gets animated later.

✓ Approve for animation — only pay for the frames you like

Each frame card has a ✓ Approved for animation tick (on by default). Untick any frame you're not happy with. When you press Generate:

• Approved frames → full AI animation (Kling/Higgsfield) — the good stuff, costs credits.
• Unticked frames → a free, simple Ken Burns zoom — so the video still assembles complete and you can see the whole cut.
• The 💰 cost panel drops as you untick frames, so you see the saving immediately.

🎬 Live clip reveal — watch it build

While a render runs, each frame's clip appears in its card the moment it's done — you don't wait for the whole video. The card shows ⏳ animating…, then flips to a looping 🎬 clip ready preview. If an early clip looks wrong you can stop, fix that frame with 🔄, untick it, and re-run — without having waited for all ten.

Timeline, safe-zone, posting kit, and export

• Timeline strip: after parsing, a horizontal strip shows frame order and duration so you can judge rhythm before rendering.
• Reels safe-zone: bottom captions are raised by default, and the output preview shows Instagram's bottom UI safe area.
• Keyword highlight: wrap a word in ==double equals==, e.g. I had ==nothing== left, to colour that phrase in captions.
• Text Card: choose Text Card on a frame, or write [layout: text_card], for a full-screen bold statement card.
• Posting kit: story mode can generate hashtags and pick a cover frame from the Caption: block. Brand copy remains operator-supplied.
• Export clips + edit list: after a completed run, download a zip with clips and edit_list.json for a human editor.
• Did this reel perform? after a completed run, the output panel has an optional views / likes / note capture — fill it in once the reel is posted and click Save. It's stored with the run as the seed of a performance-feedback loop.
• Redo motion: retry only the motion for one approved still without regenerating the image.

9 · How to build a story that lands

• 10–12 frames = a great 60–90s reel.
• Write captions in first person, present tense, punchy, under 15 words.
• Never show illness/trauma on a face — use AI Symbolic (the empty room speaks louder).
• Save big moves (360 orbit, crane up) for the single biggest moment.

10 · One-page cheat sheet (print this)

Optional: typing it in the story instead of clicking

Every button has a typed shortcut you can put under a frame line. The buttons do the same thing — use whichever you prefer.

11 · A full worked example

A real story turned into a 10-shot reel. Paste it, set the folder, hit Parse — then read why each choice was made.

Reels

Frame 1
From a farmer to a model… the story of a desi girl with big dreams.
[photo: ai_portrait]
[note: Strong proud Rajasthani woman, direct gaze, chin up. Golden hour, dust in air. A HERO, not a victim.]
[camera: crash zoom in]

Frame 5
After COVID, I was diagnosed with arthritis. I lost my hair, my confidence… and myself.
[photo: ai_symbolic]
[note: NO person. Medicine bottles on a windowsill, a hairbrush with fallen strands, cold blue-grey light. The objects carry the grief.]
[edit: cold grey winter light and frost on the window]
[camera: static]

Frame 7
Bedridden, I watched modelling videos thinking, "one day I'll walk the ramp."
[photo: lalita_face.jpg]
[lipsync: yes]
[note: The turning point — she says it herself. Phone glow on her face in a dark room. A small smile starting.]

Frame 8
I fought back — and the girl who couldn't stand walked the ramp in heels.
[photo: ramp_walk.mov]
[start: 3]
[note: Real ramp-walk clip. Head high, full confidence. The peak. Skip the first 3 seconds.]
[duration: 8]

Frame 9
I won Mrs. Rajasthan 1st runner-up.
[photo: crown.jpg]
[note: Pure victory — crown, sash, stage. She earned this. High saturation, saffron and gold.]
[camera: 360 orbit]

Caption:
From the farm to the ramp — the full story for your followers goes here.

12 · Quick fixes (FAQ)

13 · Brand / Ad Mode — full walkthrough

Click 🛍 Brand / Ad mode in the header (or go to /brand) to switch. The engine is identical — same pipeline, same frame cards, same models — but the UI gains a brand brief panel, real-asset uploads, mandatory checks, and a product beat toggle per shot.

13a · Brand brief panel

13b · Brand kit & assets (real-only — never AI-generated)

13c · Audio — VO & music

Brand mode separates the announcer voice-over from the background music. Both are decided per-project.

Need an ElevenLabs voice ID? Paste it in the VO voice field. Leave blank for the default voice.

13d · Writing the ad script & marking product beats

Write the script exactly like a story script — one caption per Frame. The caption is the on-screen text the brand supplies. After parsing, each frame card gets a 🛍 Product beat toggle.

13e · Mandatories hard-block (the checklist)

If you click Generate Ad with something missing, the tool refuses to spend any credits and shows a red checklist. You cannot generate until all boxes are ticked. This protects against accidentally publishing an ad without a logo or disclosure.

13f · Brand mode quick fixes

14 · Studio Mode — type a brief, get a reel

Open 🎬 Studio mode from the top link. Instead of writing a script, you describe the reel in plain words and the AI drafts the shots. It uses the same engine (Preview, cost, Generate, export) plus a reusable identity library.

14a · Identity library (Talent & Product)

Save a Talent (a face) and/or a Product (a photo + specs) once. Studio locks them across every shot so the same person and the same product appear consistently — and they're reusable in future reels. On shots you mark 🛍 Product beat, the real product image is used directly (never re-generated), so logos and fine detail stay exact.

14b · Brief → shots

Pick a scope — Commerce (product / fashion / jewelry ads) or General (any idea) — write your brief, and click ✨ Plan shots. The AI returns editable shot cards (on-screen line, camera move, shot size). Edit any card before Preview/Render. The AI may draft the on-screen lines here (you edit them); all text still passes the safety check.

14c · Per-shot controls

Each shot card adds a Talent/Product selector, a Negative prompt (what to avoid), and a Continuity lock (outfit/styling that must not change). Sensible defaults are prefilled; tweak per shot as needed.

Story mode workflow: Write short lines → Parse → set Picture + Note + Camera per shot → Dev test → Production final → Download.
Brand mode workflow: Fill brief → upload logo + product → write ad script → Parse → tick 🛍 product beats → Preview Stills → Generate Ad → Download.
Studio mode workflow: Save Talent/Product → pick scope → write brief → ✨ Plan shots → edit cards → Preview Stills → Generate Reel → Download.
When unsure, leave models on Auto and start in Dev.