Tangra AI Storyteller

Turn any story into a finished video - Tangra writes the scenes and creates the images, narration and music.

Overview

Tangra AI Storyteller turns a written story into a finished video through a seven-step wizard. You write or paste the story, pick the storytelling framework and the emotion the video should land, and Tangra drafts a script split into scenes with the recurring characters pulled out of it. You then set the format and the visual style, check a two-frame draft, and create every scene image, the narration voiceover and a background music track. The finished story hands off to the same timeline editor the Video Generator uses, where you trim, rearrange and export an MP4 or WebM. Every story is saved as a project in My Stories, so you can leave one half-written and come back to it.

Who it’s for

Marketers and small businesses turning a brand story, case study or launch into a video
Content creators who want a narrated, scene-by-scene video without cutting it together by hand
Educators and explainer authors turning a written piece into illustrated, narrated scenes
Authors and storytellers who want their writing on screen with consistent characters
Anyone who has the story written down and no time to film, illustrate or record it

Key features

Guided seven-step wizard

Start, Framework, Emotion, Story, Style, Create and Export. A progress bar across the top shows where you are and lets you jump back to any step you have already passed.

Nine storytelling frameworks

Pick the narrative shape: Hero's Journey, Transformation Arc, And · But · Therefore, Curiosity Loop, Before · After · Bridge, Problem · Agitate · Solve, Hook · Story · Offer, Star · Chain · Hook or Inverted Pyramid. Each card shows what it is best for, and Tangra marks the one that fits your story as Recommended.

Emotion wheel

Choose the feeling the video should land from a wheel of eight families (Happy, Loved, Confident, Playful, Embarrassed, Angry, Scared, Sad), each with three finer shades. Tangra reads an emotion out of your story and pre-selects it, and the choice also suggests narration voices that carry that feeling.

AI-drafted script with characters

Tangra splits the story into scenes, each with its own narration and visual prompt, and pulls out the recurring characters. Rewrite any scene, drag scenes into a new order, insert or delete scenes, and edit a character's description or attach a reference image so they look the same throughout.

Look & feel art direction

Describe the visual style in your own words, or start from a preset - Cinematic warm, Cartoonish & vibrant, Anime brush strokes, Subdued → bright, Minimal & clean or Documentary. Tangra proposes the format and the style from your story before you change anything.

Two-frame draft before you commit

Create Draft renders two sample frames in your style so you can check the look and re-create them until you are happy, before creating the whole story.

Model quality tiers

Fast, Lite, Pro and 4K each pair a scriptwriting model with an image model. The Start step quotes what a story costs in credits on the tier you pick, and how many stories your balance covers.

Scene grid with per-scene control

Create All Media renders every missing image and narration in one run, or work scene by scene: swap the image model, re-create a single image, edit and re-voice the narration, or pick a voiceover from your library. Switch between a detailed and a compact grid.

Animate a scene

Turn a still into motion with a free camera effect (Zoom In/Out, Ken Burns, Pan, Slow Rotate, Handheld, Color Pulse, Dynamic) applied to one scene or all of them, or run an AI video model on it - including avatar models that lipsync the scene's narration.

Narration and background music

Narration uses the same voice library as the Voice Generator, with voices recommended for the emotion you chose. Background music is composed from a mood you describe or picked from your library. Either can be switched off.

My Stories

Every story is saved automatically as a project. Search your stories, rename, duplicate or delete them, and reopen one on the step you left it.

Timeline editor and export

Export Video builds the timeline and starts the render straight away; Edit Timeline opens the same multi-track editor as the Video Generator first, so you can trim, reorder or add clips, music and text. An optional watermark can be placed over every scene.

How to use it — step by step

Open Storyteller
Go to https://tangra.link/storyteller or open the app. With no stories yet you land straight in the wizard; once you have some, you land on My Stories and start a new one with New Story.
Write or paste your story
Type the story into the box on the Start step, or load one of the example ideas. At least 20 characters are needed to continue. Pick the model quality here too - the step quotes roughly what the story will cost.
Pick a storytelling framework
Choose the narrative shape from the nine framework cards, or leave the 'Not sure?' card selected and let Tangra pick the one that fits. The card Tangra recommends is marked.
Dial in the emotion
Pick the feeling the video should land on the emotion wheel, or keep the emotion Tangra read out of your story.
Review the story and characters
Tangra drafts the script. Edit any scene's narration and visual prompt, drag scenes to reorder them, add or delete scenes, and go through the recurring characters - edit their descriptions and attach a reference image so they stay consistent.
Set the visual style
Choose the format (aspect ratio and size), describe the look & feel or pick a preset, then click Create Draft to render two sample frames. Re-create them until the style is right.
Create the scenes
Turn narration and background music on or off, pick the narration voice and describe the music mood, then click Create All Media. Tangra renders every scene image and voiceover; you can also re-create a single scene or animate it.
Export your video
On the Export step add a watermark if you want one, then click Export Video to build the timeline and start the render, or Edit Timeline to open the editor first and trim, reorder or add to the story before exporting.

See it in action

Screens you’ll work with inside Storyteller:

My Stories — The Storyteller home screen listing saved stories, with the search box, the New Story button and rename / duplicate / delete on each card. — Fig 1. My Stories

Start step — The wizard's first step with the story box, the example ideas, the model quality picker and the credit estimate. — Fig 2. Start step

Framework step — The nine storytelling framework cards with their diagrams, the Recommended pill on Tangra's pick, and the 'Not sure?' card. — Fig 3. Framework step

Emotion wheel — The emotion wheel with its eight families and their shades, next to the detail card for the selected feeling. — Fig 4. Emotion wheel

Story step — The drafted script with one row per scene, above the recurring characters and their reference images. — Fig 5. Story step

Style step — The format button, the look & feel box with its presets, and the two draft frames rendered in that style. — Fig 6. Style step

Create step — The scene grid with images and narration, the background music and voice panels above it, and the Create All Media button. — Fig 7. Create step

Export step — The export step with the watermark panel open and the Export Video / Edit Timeline buttons. — Fig 8. Export step

Settings & controls

Every control in Storyteller, what it does and the options available:

Control	What it does	Options
Your story	The text Tangra turns into scenes.	Free text, at least 20 characters; six example ideas to load
Model quality	Picks the model that writes the script and the model that creates the scene images, and so sets the credit cost.	Fast, Lite, Pro, 4K
Storytelling framework	The narrative shape the script follows.	Hero's Journey, Transformation Arc, And · But · Therefore, Curiosity Loop, Before · After · Bridge, Problem · Agitate · Solve, Hook · Story · Offer, Star · Chain · Hook, Inverted Pyramid, or 'Not sure?' to let Tangra pick
Emotion wheel	The feeling the video should land; also drives the recommended narration voices.	Happy, Loved, Confident, Playful, Embarrassed, Angry, Scared and Sad, each with three shades - or leave it on Tangra's reading
Scene list	The drafted script, one row per scene with its narration and its visual prompt.	Edit, drag to reorder, add a start / end / in-between scene, delete
Characters & reference images	The recurring characters pulled out of the script, kept looking the same across scenes.	Name, description and a reference image (uploaded or picked from your library); add or remove characters. Reference images are available on the Pro and 4K tiers
Format	Aspect ratio and pixel size every scene image is rendered at.	16:9, 9:16, 1:1, 4:5, 4:3, 2:3, 21:9 and 2.39:1 cinema, in sizes from SD up to 4K, plus social presets
Look & feel	The art direction applied to every scene image.	Free text, or the presets Cinematic warm, Cartoonish & vibrant, Anime brush strokes, Subdued → bright, Minimal & clean, Documentary
Create Draft	Renders two sample frames so you can check the style before creating the rest.	Each frame can be re-created on its own
Background music	One instrumental track under the whole story.	On / off; describe the mood and create a track, or select one from your library
Voice narration	Reads each scene's narration.	On / off; any voice from the Tangra voice library, with voices recommended for the emotion you picked
Image model (Create step)	The model that renders the scene images; starts on the model your tier selected.	Any image model that supports the chosen size, each showing its credit cost
Create All Media	Renders every missing scene image and narration, and the music track, in one run.	Becomes 'Create Remaining' once part of it exists; can be cancelled while it runs
Scene view	How the scene grid is laid out.	Detailed, Compact
Animate	Turns a scene still into motion.	Effects (Zoom In/Out, Ken Burns, Pan Left/Right/Up/Down, Slow Rotate, Handheld, Color Pulse, Dynamic) applied to one scene or all of them; or AI video from a motion prompt; or an avatar model that lipsyncs the narration
Watermark	A logo laid over every scene.	Optional; upload or select a logo and choose where it sits, with a live preview
Export Video / Edit Timeline	Hands the finished story to the timeline editor.	Export Video starts the render on arrival; Edit Timeline opens the editor first

AI models

Storyteller lets you pick from a curated lineup of AI models — each shows its credit cost in the app. Available models:

Fast tier (script: ChatGPT Luna, images: Fast)Lite tier (script: Claude Haiku, images: Lite)Pro tier (script: ChatGPT Pro, images: References)4K tier (script: Claude Opus, images: Nano Banana 2)Any image model from the picker on the Create stepMusic 4.0 (instrumental background track)Speakers (voice1) for the narration voicesAnimate: Horizon, Cinema, Lite, Pro, Studio, Fast, Pro 2, 4KAnimate with lipsync: Portrait, Studio Cast, Director, Speaker, Speaker Pro, Performer, Performer Pro, Influencer, Influencer Fast, Narrator, Presenter

Outputs & formats

MP4 (H.264 / H.265) and WebM (VP9), exported from the timeline editor
Scene images at the format you pick: 16:9, 9:16, 1:1, 4:5, 4:3, 2:3, 21:9 or 2.39:1 cinema, in sizes from SD up to 4K
One narration voiceover per scene
One instrumental background music track
Animated scenes as short motion or lipsync clips
The whole story kept as a saved project you can reopen, edit and export again

Pro tips

Paste the story you already have - Storyteller is built around a written story, not a one-line prompt. The more the story says, the better the framework, the emotion and the scene split.

Leave the framework and the emotion on Tangra's recommendation the first time round; it reads both out of your story, and its picks are usually the right shape.

Always run Create Draft before creating the rest. Two frames cost a fraction of a whole story, and checking the look & feel wording is exactly what they are for.

Give recurring characters a description and a reference image on the Story step (Pro and 4K tiers) - that is what keeps a face the same from scene to scene.

Draft on Fast or Lite and switch tiers once the story and the style are settled: the tier drives both the script model and the image model, so it drives the cost.

Pick the format before you create the scenes. Every scene image is rendered at that aspect ratio, so changing it later means creating them again.

Choose Edit Timeline instead of Export Video when you want to trim, add a clip or drop in text - it is the same editor as the Video Generator, and you export from there.

Storyteller FAQ

What is Tangra AI Storyteller?

Storyteller is a guided wizard that turns a story you have written into a finished video. It picks a storytelling framework and an emotion, drafts a script split into scenes, creates an image for each scene, narrates them and composes background music, then hands the result to the timeline editor to export.

How is Storyteller different from the Video Generator?

The Video Generator creates individual clips from a prompt or an image, which you then assemble yourself. Storyteller starts from a written story and produces the whole thing - scenes, images, narration and music - in one guided flow. Both finish in the same timeline editor.

How many scenes will my story have?

Tangra decides: it splits the story into as many scenes as it needs, so there is no scene-count setting. The estimate on the first step is quoted off a typical six-scene story. On the Story step you can add or delete scenes yourself.

Can I keep a character looking the same in every scene?

Yes. Tangra pulls the recurring characters out of your script on the Story step. Edit each one's description and attach a reference image - your own upload or a picture from your library - and Storyteller uses it when it renders the scenes. Reference images are available on the Pro and 4K tiers.

Can I change the images, narration or music afterwards?

Yes. On the Create step any scene can be re-created with a different image model, its narration edited and re-voiced, or a voiceover picked from your library. The music track can be composed again from a new mood description or replaced from your library, and narration and music can each be switched off.

Can I make the scenes move?

Yes. Animate a scene with a free camera effect - zoom, pan, Ken Burns, slow rotate, handheld, color pulse - which you can apply to every scene at once, or run an AI video model on the still. Avatar models go further and lipsync the scene's narration.

Where are my stories saved?

Every story is saved automatically as a project. My Stories lists them all, and you can search, rename, duplicate or delete a story and reopen it on the step you left it.

What do I get at the end?

The finished story opens in the timeline editor, where you export it as an MP4 or WebM at the resolution and quality you choose. You can also open the timeline first and trim, reorder or add clips, music and text before exporting.

How much does a story cost?

A story spends credits on the script, one image per scene, one narration per scene and one music track. The Start step quotes the whole story for the tier you pick, and the Create step quotes what is still missing before you press the button.