The Formula
Shot + Scene + Action + Character + Camera + Audio, written as one flowing paragraph in present tense.
- Shot
- A cinematography term and shot scale that fit the genre, such as a slow push-in close-up.
- Scene
- Lighting, colour palette, surface textures and atmosphere.
- Action
- The core action as a natural sequence from beginning to end.
- Character
- Age, hair, clothing and distinguishing features, with emotion shown through physical cues.
- Camera
- How and when the camera moves, described relative to the subject, and how the subject looks after the move.
- Audio
- Ambient sound, music and speech, with dialogue in quotation marks.
The Rules
Write one paragraph in present tense
LTX has no bracket or tag syntax. Everything goes in a single flowing paragraph of present-tense prose.
Show emotion through the body
Write "jaw tight, eyes not leaving the door", not "nervous" or "sad". The official guide makes this point twice.
Keep one light logic
Use one coherent lighting setup per shot. Mixed light sources confuse the result.
Aim for 4 to 8 sentences
Match detail to shot scale: close-ups need more description than wide shots. Every sentence should add something the viewer sees or hears.
Describe the camera relative to the subject
Say how and when the camera moves and how the subject appears after the move. That helps the model complete the motion.
Quote dialogue and name the accent
Put spoken words in quotation marks and say the language and accent, for example: She whispers in a London accent, "He's late."
Stay in one continuous take from an image
When animating a start frame, describe one unbroken take. Never write a shot list, numbered beats or screenplay sluglines.
Keep on-screen text and wild physics out
Spelling across frames is not guaranteed, so add titles in your edit. Simple, plausible motion holds up better than chaotic physics.
A Template You Can Copy
A [shot scale and camera term] [reveals or frames] [character: age, hair, clothing, distinguishing feature] in [scene: light source, palette, textures, atmosphere]. [Action in present tense, as a natural sequence, with emotion shown through physical cues]. The camera [move] [relative to the subject], and as it settles [how the subject now appears]. [Speaker] says in a [accent] accent, "[exact words]". [Ambient sound, music].
A Real Example
The shot: Night, rain-streaked diner window. A woman in a red coat sits alone in the last booth, turning a ring on the table. Slow push in from the window. Mara says: "You kept it." (6 seconds, 16:9)
A slow push-in through a rain-streaked diner window at night reveals a woman in a red coat sitting alone in the last booth, warm interior light glowing against the cold blue dark outside. Water runs in thin rivulets down the glass, softly distorting her silhouette as the camera gradually moves closer. She sits with her shoulders curled inward, slowly turning a ring between her thumb and forefinger on the tabletop, her gaze fixed downward, jaw tight. Her dark hair falls loosely around her face, the collar of her red coat damp at the edges. As the camera settles closer, her expression comes into focus, quiet and unmoving, before she lifts her eyes. She says in a soft, flat American accent, "You kept it." The hush of rain tapping the window blends with the low murmur of an empty diner and the faint hum of a distant neon sign.
It is one present-tense paragraph that moves through shot, scene, character, action, camera and audio without any tags. Emotion is physical (shoulders curled inward, jaw tight, gaze fixed downward), the camera move is described relative to her and what it reveals as it settles, and the line is quoted with an accent named.
Dialogue
Put the exact words in quotation marks inside the paragraph, and name the language and accent with the delivery. Volume words the guide lists are whisper, mutter, shout and scream. Write the line in the native script of its language, because the model does not translate for you. LTX 2.3 does not speak, so leave lines out for it.
She whispers in a London accent, "He's late."
Camera Moves
| Move | How to write it for LTX 2.5 |
|---|---|
| Slow push in | The camera slowly pushes in until her face fills the frame and her eyes lift. |
| Pull back | The camera pulls back to reveal the empty room around him. |
| Follow | The camera follows her from behind as she walks into the crowd. |
| Orbit | The camera circles around the table, ending on his face. |
| Tilt up | The camera tilts upward from her boots to the neon sign above her. |
| Static | A static frame holds on the doorway as he steps through. |
Common Mistakes
| Mistake | Fix |
|---|---|
| Abstract emotion labels such as "she looks sad". | Describe the body: "her shoulders curl inward, gaze fixed on the table". |
| Writing a shot list or numbered beats. | Write one continuous paragraph. If you really need a cut, name it in prose ("A hard cut transitions to..."). |
| Mixing several light sources with no logic. | Pick one lighting setup, such as warm interior light against a cold night outside, and keep it. |
| Asking for titles or signs with exact spelling. | Keep text short and prominent, or add it in your edit. |
| Borrowing brackets or tags from other models. | LTX reads plain prose; put speech in ordinary quotation marks. |
Length, Resolution and Price
LTX 2.5 Pro makes 6, 8 or 10 second clips at $0.12 per second at 720p or $0.17 at 1080p, and LTX 2.5 Fast runs 6 to 20 seconds in 2 second steps at $0.09 (720p), $0.13 (1080p), $0.19 (1440p) or $0.30 (4K) per second. Both speak with optional sound and accept an end frame, in 16:9 or 9:16. The older LTX 2.3 Pro ($0.08 per second) and LTX 2.3 Fast ($0.06 per second) render 6, 8 or 10 seconds at 1080p and do not speak (checked 2026-10-22). Every model side by side is on the model spec sheet.