The Complete Guide to Prompting OpenAI Sora 2: Tips, Techniques & Templates
A comprehensive, hands-on guide to writing effective prompts for OpenAI's Sora 2 video generation model. Includes copyable prompt templates for cinematic scenes, dialogue, image input, remixing, and more.
OpenAI's Sora 2 is a state-of-the-art AI video generation model that turns text — and images — into high-fidelity video clips. In this guide, we break down everything you need to know to prompt Sora 2 effectively, from API parameters to cinematic prompt structure, with ready-to-copy templates you can use right away.
What is Sora 2?
Sora 2 is OpenAI's next-generation text-to-video model, accessible via the API and Sora.com. It supports both sora-2 and the more capable sora-2-pro tiers. Key highlights include:
- High-Fidelity Video: Generates realistic, cinematically rich video from text descriptions.
- Image Input: Use a reference image as the first frame for finer compositional control.
- Dialogue & Audio: Describe spoken lines and soundscapes directly in your prompt.
- Remix Functionality: Iteratively tweak existing generations for controlled refinement.
- Multiple Resolutions & Lengths: Generate clips from 4 to 12 seconds in HD resolutions.
API Parameters
Before writing your prompt, understand that some video attributes are set only via API parameters — you cannot request them in prose:
| Parameter | Options |
|---|---|
model | sora-2 or sora-2-pro |
size | 1280x720, 720x1280 (sora-2); additionally 1024x1792, 1792x1024 (sora-2-pro) |
seconds | "4" (default), "8", "12" |
Tip: For best results, aim for concise 4-second clips. If you need more time, stitching two 4-second clips together in editing often yields better quality than a single longer generation.
The Philosophy of Prompting Sora 2
Think of prompting like briefing a cinematographer who has never seen your storyboard. If you leave out details, they'll improvise — and you may not get what you envisioned.
At the same time, leaving some details open can be powerful. Giving the model more creative freedom leads to surprising, beautiful variations. Detailed prompts give you control and consistency; lighter prompts open space for creative outcomes.
Key principles:
- Iterate: Small changes to camera, lighting, or action can shift the outcome dramatically.
- Variety: Using the same prompt multiple times gives different results — this is intentional.
- Collaborate: You provide the direction, the model delivers creative variations.
Prompt Anatomy That Works
A clear prompt describes a shot as if you were sketching it onto a storyboard. Include:
| Element | What It Controls | Example |
|---|---|---|
| Visual Style | Overall aesthetic, genre, medium | "1970s romantic drama, shot on 35mm film", "Stop-motion animation" |
| Camera / Framing | Shot type, angle, depth of field | "Medium close-up, slow push-in", "Wide establishing shot, eye level" |
| Lighting & Color | Mood, atmosphere, palette | "Warm tungsten key, cool rim from window", "Golden hour backlight" |
| Characters | Appearance, costume, expression | "A mid-30s traveler in a navy coat" |
| Environment | Setting, scenery, props | "A cluttered workshop with shelves of gears and blueprints" |
| Actions | Scene events, beats, gestures | "Actor takes four steps to the window, pauses, pulls the curtain" |
| Dialogue | Spoken lines | Detective: "You're lying. I can hear it in your silence." |
| Sound | Ambient noise, music, effects | "Rain, ticking clock, soft mechanical hum" |
Prompting Techniques
1. Start with Visual Style
Establishing the aesthetic early sets a visual tone that the model carries through consistently. The same subject will look completely different depending on whether you call for a "polished Hollywood drama," a "handheld smartphone clip," or a "grainy vintage commercial."
❌ Weak:
Camera shot: cinematic look✅ Strong:
Camera shot: wide shot, low angle
Depth of field: shallow (sharp on subject, blurred background)
Lighting + palette: warm backlight with soft rim2. Control Motion with Precision
Movement is often the hardest element to get right. Keep it simple: one clear camera move and one clear subject action per shot. Describe actions in beats — small steps, gestures, or pauses — so they feel grounded in time.
❌ Weak:
Actor walks across the room.✅ Strong:
Actor takes four steps to the window, pauses, and pulls the curtain in the final second.Some examples for good camera instructions:
wide establishing shot, eye levelwide shot, tracking left to rightaerial wide shot, slight downward anglemedium close-up, slow push-in with gentle parallax
3. Be Specific with Lighting & Color
Light determines mood as much as action or setting. When cutting multiple clips together, consistent lighting logic makes the edit seamless.
❌ Weak:
Lighting + palette: brightly lit room✅ Strong:
Lighting + palette: soft window light with warm lamp fill, cool rim from hallway
Palette anchors: amber, cream, walnut brownDescribe three to five palette anchors to keep colors stable across shots.
4. Write Dialogue Naturally
Place dialogue in a dedicated block below your scene description so the model clearly separates visual from spoken content. Keep lines concise and natural:
- A 4-second clip accommodates one or two short exchanges.
- An 8-second clip can support a few more lines.
Label speakers consistently in multi-character scenes to help the model match lines to expressions and gestures.
A cramped, windowless room with walls the color of old ash. A single bare bulb dangles from the ceiling. Two chairs face each other across a scarred metal table. The Detective sits with eyes sharp and unblinking. Across from him, the Suspect slouches, cigarette smoke curling lazily toward the ceiling.
Dialogue:
- Detective: "You're lying. I can hear it in your silence."
- Suspect: "Or maybe I'm just tired of talking."
- Detective: "Either way, you'll talk before the night's over."5. Use Image Input for Compositional Control
For even more precision, attach a reference image to your API call using the input_reference parameter. The model uses the image as an anchor for the first frame, while your text prompt defines what happens next.
Requirements:
- The image must match the target video's resolution (
size). - Supported formats:
image/jpeg,image/png,image/webp.
Example prompts with image input:
"She turns around and smiles, then slowly walks out of the frame.""The fridge door opens. A cute, chubby purple monster comes out of it."Experimentation tip: Don't have a reference image? Use OpenAI's image generation model to quickly produce environments and scene designs, then pass them into Sora as references.
6. Iterate with the Remix Feature
Remix is for nudging, not gambling. Make one controlled change at a time and describe exactly what you're changing:
"Same shot, switch to 85mm""Same lighting, new palette: teal, sand, rust"
If a shot keeps misfiring, strip it back: freeze the camera, simplify the action, clear the background. Once it works, add complexity step by step.
7. Go Ultra-Detailed for Cinematic Shots
For complex, professional-grade shots, you can describe the look, camera setup, grading, soundscape, and shot rationale in production terms — similar to how a director briefs a camera crew.
Format & Look
Duration 4s; 180° shutter; digital capture emulating 65mm photochemical contrast; fine grain; subtle halation on speculars.
Lenses & Filtration
32mm / 50mm spherical primes; Black Pro-Mist 1/4; slight CPL rotation to manage glass reflections.
Grade / Palette
Highlights: clean morning sunlight with amber lift. Mids: balanced neutrals with slight teal cast in shadows. Blacks: soft, neutral with mild lift for haze retention.
Lighting & Atmosphere
Natural sunlight from camera left, low angle (07:30 AM). Gentle mist; train exhaust drift through light beam.
Location & Framing
Urban commuter platform, dawn. Foreground: yellow safety line, coffee cup on bench. Midground: waiting passengers silhouetted in haze. Background: arriving train braking to a stop.
Sound
Diegetic only: faint rail screech, train brakes hiss, distant announcement muffled (-20 LUFS). No score or added foley.
Shot List (4s total)
0.00–2.40 — "Arrival Drift" (32mm, shoulder-mounted slow dolly left): Camera slides past platform signage; shallow focus reveals traveler looking down tracks.
2.40–4.00 — "Turn and Pause" (50mm, slow arc in): Traveler turns toward camera, catching sunlight rim across cheek.Prompt Structure Template
Use this framework as a starting point. Not every section is required — leave out anything that doesn't matter for your shot:
[Prose scene description: characters, costumes, scenery, weather, and other details]
Cinematography:
Camera shot: [framing and angle, e.g. wide establishing shot, eye level]
Mood: [overall tone, e.g. cinematic and tense, playful and suspenseful]
Actions:
- [Action 1: a clear, specific beat or gesture]
- [Action 2: another distinct beat within the clip]
- [Action 3: another action or dialogue line]
Dialogue:
[Short, natural lines here. Keep them brief so they match the clip length.]
Background Sound:
[Ambient noise, music cues, or sound effects]Ready-to-Copy Prompt Examples
🤖 Animated Short (Whimsical / Stop-Motion Feel)
Style: Hand-painted 2D/3D hybrid animation with soft brush textures, warm tungsten lighting, and a tactile, stop-motion feel. The aesthetic evokes mid-2000s storybook animation — cozy, imperfect, full of mechanical charm.
Inside a cluttered workshop, shelves overflow with gears, bolts, and yellowing blueprints. At the center, a small round robot sits on a wooden bench, its dented body patched with mismatched plates. Its large glowing eyes flicker pale blue as it fiddles nervously with a humming light bulb.
Cinematography:
Camera: medium close-up, slow push-in with gentle parallax from hanging tools
Lens: 35mm virtual lens; shallow depth of field
Lighting: warm key from overhead practical; cool spill from window
Mood: gentle, whimsical, a touch of suspense
Actions:
- The robot taps the bulb; sparks crackle.
- It flinches, dropping the bulb, eyes widening.
- The bulb tumbles in slow motion; it catches it just in time.
- A puff of steam escapes its chest — relief and pride.
- Robot says quietly: "Almost lost it… but I got it!"
Background Sound: Rain, ticking clock, soft mechanical hum, faint bulb sizzle.💃 1970s Romantic Drama (Rooftop Dance)
Style: 1970s romantic drama, shot on 35mm film with natural flares, soft focus, and warm halation. Slight gate weave and handheld micro-shake evoke vintage intimacy. Warm Kodak-inspired grade; film grain and soft vignette for period authenticity.
At golden hour, a brick tenement rooftop transforms into a small stage. Laundry lines strung with white sheets sway in the wind. Strings of mismatched fairy bulbs hum overhead. A young woman in a flowing red silk dress dances barefoot, curls glowing in the fading light. Her partner — sleeves rolled, suspenders loose — claps along, his smile wide and unguarded.
Cinematography:
Camera: medium-wide shot, slow dolly-in from eye level
Lens: 40mm spherical; shallow focus to isolate the couple from skyline
Lighting: golden natural key with tungsten bounce; edge from fairy bulbs
Mood: nostalgic, tender, cinematic
Actions:
- She spins; her dress flares, catching sunlight.
- Woman (laughing): "See? Even the city dances with us tonight."
- He steps in, catches her hand, and dips her into shadow.
- Man (smiling): "Only because you lead."
- Sheets drift across frame, briefly veiling the skyline before parting again.
Background Sound: Natural ambience only: faint wind, fabric flutter, street noise, muffled music. No added score.🕵️ Crime Interrogation Scene
A cramped, windowless room with walls the color of old ash. A single bare bulb dangles from the ceiling, its light pooling onto the scarred metal table at the center. Two chairs face each other across it. The Detective sits, trench coat draped across the back of his chair, eyes sharp and unblinking. Across from him, the Suspect slouches, cigarette smoke curling lazily toward the ceiling.
Dialogue:
- Detective: "You're lying. I can hear it in your silence."
- Suspect: "Or maybe I'm just tired of talking."
- Detective: "Either way, you'll talk before the night's over."🚉 Cinematic Urban Platform (Ultra-Detailed)
Format & Look: Duration 4s; 180° shutter; digital capture emulating 65mm photochemical contrast; fine grain; subtle halation on speculars; no gate weave.
Lenses & Filtration: 32mm / 50mm spherical primes; Black Pro-Mist 1/4; slight CPL rotation to manage glass reflections on train windows.
Grade / Palette: Highlights: clean morning sunlight with amber lift. Mids: balanced neutrals with slight teal cast in shadows. Blacks: soft, neutral with mild lift for haze retention.
Lighting & Atmosphere: Natural sunlight from camera left, low angle (07:30 AM). Gentle mist; train exhaust drift through light beam.
Location & Framing: Urban commuter platform, dawn. Foreground: yellow safety line, coffee cup on bench. Midground: waiting passengers silhouetted in haze. Background: arriving train braking to a stop.
Sound: Diegetic only: faint rail screech, train brakes hiss, distant announcement muffled (-20 LUFS), low ambient hum. No score or added foley.
0.00–2.40 — "Arrival Drift" (32mm, shoulder-mounted slow dolly left): Camera slides past platform signage edge; shallow focus reveals traveler mid-frame looking down tracks. Morning light blooms across lens.
2.40–4.00 — "Turn and Pause" (50mm, slow arc in): Traveler turns slightly toward camera, catching sunlight rim across cheek.🌿 Short & Creative (Open-Ended)
In a 90s documentary-style interview, an old Swedish man sits in a study and says, "I still remember when I was young."Short prompts like this give the model creative freedom over character look, set design, camera angle, and lighting. Great for exploring unexpected interpretations.
Pro Tips for Better Results
- Set style first: Begin with the visual medium or film aesthetic — this frames all other choices.
- One action, one camera move per shot: Keep it simple; complexity compounds unpredictably.
- Use filmmaking terminology: "Medium close-up", "dolly-in", "handheld", "rack focus" give the model precise direction.
- Name palette anchors: Three to five color names keep your palette stable across shots.
- Keep dialogue brief: Match line count to clip length — don't pack an 8-line exchange into a 4-second clip.
- Remix with intention: Change one variable at a time and describe the change explicitly.
- Use image references: Generate a reference image first, then use it as a Sora input for locked composition.
- Embrace variation: Re-running the same prompt yields different results. Treat each generation as a fresh creative take.
Conclusion
Sora 2 rewards precision and experimentation in equal measure. The key to getting great results is writing structured, cinematic prompts — specifying style, camera work, characters, actions, dialogue, and sound — while leaving room for the model to breathe and surprise you.
Whether you're a filmmaker, content creator, or developer, mastering Sora 2 prompting opens up powerful new creative possibilities. Start with the ready-to-copy examples above, experiment with image input and remix, and iterate your way to the shot you envision.
Ready to start? Try Sora 2 now at sora.com or via the OpenAI API.
Author
Categories
More Posts
The Guide to Prompting Kling 3.0: Tips, Techniques & Prompt Templates
A comprehensive, hands-on guide to writing effective prompts for Kuaishou's Kling 3.0 AI video generation model. Includes copyable prompt templates for multi-shot storytelling, dialogue, camera work, audio design, and more.
AI Video Prompts: The Complete Guide to Cinematic, Viral & Movie-Quality Results
Master AI video prompts for every use case — from Sora 2 video prompts and lifelike Veo 3 prompts to prompts for movie-quality video generation and viral YouTube content. Includes copyable examples for every major model.
Top 10 Best AI Video Generators in 2026
We personally tested the top 10 AI video generators in 2026 using the same prompt. Here's how Runway, Kling AI, OpenAI Sora, Google Veo 3, Synthesia, HeyGen, Pika, Luma, Adobe Firefly, and Manus actually performed.
Waitlist
Early Access
Be the first to know when AcceptPrompt launches. Sign up to get early access and exclusive updates.
Be the first to join. Free early access, 50% off when subscribe. No spam, ever.