Description
Sora (OpenAI)
What is Sora?
Versions and Models
Sora 2 (September 2025)
- Synchronized audio: Dialogues, sound effects, ambient noise
- Enhanced physics: Basketball rebounds realistically, objects persist
- Advanced world simulation: Models physical world better
- Improved controllability: Follows intricate multi-shot instructions
- Realistic styles: Cinematic, anime, realistic rendering
- Audio generation: Dialogues synchronized with lip movements
- Higher quality experimental model
- Only for ChatGPT Pro ($200/month)
- Better resolution and duration
- Unlimited relaxed generations (after 500 priority)
Sora 1.0 Turbo (December 2024)
- Much faster than Feb 2024 preview
- Still limited by physics/complexity
- Available via API
- Maintains access for existing users
Original Sora (Preview February 2024)
- Initial "jaw-dropping" demos
- Limited red team access
- GPT-1 moment for video
- Object permanence emerged
Key Features
Text-to-Video
- Generate videos from text descriptions
- Multiple styles: cinematic, realistic, anime
- Vertical short videos (social media optimized)
- Duration: up to ~20-30 seconds (unofficial)
Image-to-Video
- Animate static images
- Maintain visual consistency
- Natural motion sequences
- Concept art to motion
Synchronized Audio
- Dialogue: Voices synchronized with lip movements
- Sound effects: Aligned with on-screen action
- Ambient noise: Realistic background soundscapes
- High degree of realism: Coherent audio
Advanced Physics Simulation
- Basketball bounces off backboard if missed
- Models failure, not just success
- Improved object permanence
- Realistic motion and interactions
Multi-Shot Consistency
- Follows instructions spanning multiple shots
- Accurately persists world state
- Better continuity vs Sora 1
- Limitation: Long-form storytelling still challenging
Cameos Feature
- Upload yourself/others into AI videos
- Consent-based: only you decide who uses your likeness
- Revoke access anytime
- View all videos with your character
- Works for humans, animals, objects
Controllability
- Intricate multi-shot instructions
- Controllable camera movements
- Style adjustments
- Scene composition control
- Limitation: Prompt adherence not perfect
Social Features (Sora App)
- Feed-like functionality
- Share AI-generated videos
- Community platform
- "SlopTok" nickname by some users
- Parental controls available
Safety & Provenance
- Visible watermark: Moving digital watermark (though removable by 3rd-party tools)
- C2PA Content Credentials: Embedded provenance
- Multi-modal moderation: Input prompts, output frames, audio, scenes
- Stricter teen limits: Daily generation caps
- Character consent: Explicit permission required
Pricing
ChatGPT Plus ($20/month)
- 50 videos/month at 480p
- Or fewer videos at 720p
- Sora 2 included at no additional cost
- Priority access over free tier
- Unlimited in sense of no hard cap, subject to moderation
ChatGPT Pro ($200/month)
- 500 priority videos/month
- Sora 2 Pro model access (higher quality)
- Unlimited relaxed generations after 500 priority
- Higher resolutions
- Longer durations
- Skip waitlist (invites)
- 10x more usage vs Plus
Free Tier (Sora 2)
- Invite-only initially
- Generous limits but compute-constrained
- Available in US/Canada first
- iOS app (Android pending)
- Web access at sora.com after invite
- Future: OpenAI plans option to pay for extra videos
API Pricing (Planned)
- Sora 1.0 Turbo: Already available in API
- Sora 2 API: Planned, timeline TBD
- Unofficial providers: $0.10-0.50/second (official) vs $0.015-0.10 (3rd-party)
- ~$1-5 per 10-second video (official)
Limitations and Controversies
Technical Limitations
❌ Complex actions: Struggles with complex long duration actions
❌ Long-form consistency: Long multi-shot narratives difficult
❌ Prompt adherence: "More controllable" ≠ perfect
❌ Undocumented specs: Duration/resolution/fps not officially documented
❌ Compute intensive: "Much, much more expensive" than text/image
Controversies
- Uses copyrighted material by default unless opt-out
- Disney deal $1B (Dec 2025): 200+ licensed characters
- Japan's Content Overseas: demands stop (Ghibli, Square Enix)
- MPA criticized approach (Oct 2025)
- "Granular control" promised for copyright holders
- 3rd-party tools removed watermark 7 days after launch
- Undermines safety measures
- Nov 2024: API key leaked by testers
- Manifesto: protests "art washing"
- OpenAI revoked access 3 hours later
- Hank Green and others: app is AI slop
- Wired: overly similar to TikTok
- Concerns: misinformation, disinformation, scams
Access Restrictions
❌ No Android: Early phase
❌ Age 18+: Not available to minors
❌ Geo-restricted: No UK, Switzerland, EEA
❌ No Team/Enterprise/Edu: Only Plus/Pro/Business
Safety Restrictions
❌ People uploads limited: Deepfake mitigations
❌ Cameos: Explicit consent required
❌ Refusals: Multi-stage safety checks may reject
Use Cases
- TikTok, Reels, YouTube Shorts
- Vertical short-form content
- Viral creative videos
- Community sharing
- Concept reels
- Stylized shorts
- Pre-visualization
- Mood boards
- Product teasers
- Brand snippets
- Campaign visuals
- Explainer videos
- Rapid concepting
- Storyboarding
- Visual prototypes
- Pre-vis workflows
- Lesson visuals
- Educational explainers
- Tutorial content
- Illustrative reports
- Short films
- Creative animations
- Character-driven bits
- Dialogue-led content
- Client presentations
- Pitch decks
- Concept testing
- Iterative animation
Advantages
✅ Synchronized audio: Unique with native dialogue + sound effects
✅ Advanced physics: Better world simulation than competitors
✅ ChatGPT integration: Unique ecosystem
✅ Cameos: Upload yourself/friends with consent
✅ Multi-shot control: Persist world state across shots
✅ Social app: Built-in distribution platform
✅ Safety-first: Provenance, watermarks, moderation
✅ Pro unlimited: Unlimited relaxed generations (Pro plan)
✅ API coming: Developer access planned
Comparison vs Competitors
- Sora 2: Better audio sync, OpenAI ecosystem
- Runway: More editing tools, established
- Sora 2: Better controllability
- Veo 3: Polished lip-sync, integrated audio
- Sora 2: Superior physics, audio, realism
- Pika: More accessible, user-friendly, no waitlist
- Sora 2: Audio generation, multi-shot
- Luma: Human motion quality in certain domains
Company
Founded: 2015
Sora Launch: Preview Feb 2024 → Public Dec 2024 → Sora 2 Sep 2025
Access: ChatGPT Plus/Pro/Business
Regions: US, Canada (expanding)
Platforms: iOS app, Web (sora.com), API (planned)
- Diffusion transformer architecture
- Adaptation of DALL-E 3 tech
- Denoising latent diffusion model
- Transformer denoiser
- 3D patches in latent space
Key Features
Sora 2: synchronized audio (dialogue, sound effects, ambient)
Advanced physics simulation: objects persist, basketball bounces
Multi-shot consistency: persist world state across shots
Cameos: upload yourself/friends with consent control
Text-to-video: cinematic, realistic, anime styles
Image-to-video: animate static images with natural motion
Sora 2 Pro: higher quality, unlimited relaxed gens (Pro)
TikTok-style app: social feed, community sharing
Visible watermark + C2PA Content Credentials
Multi-modal moderation: prompts, frames, audio, scenes
Controllability: camera movements, style, scene composition
ChatGPT Plus: 50 videos/month 480p ($20/month)
ChatGPT Pro: 500 priority + unlimited relaxed ($200/month)
iOS app + web access (sora.com)
Invite-only rollout (US/Canada first)
API planned (Sora 1.0 Turbo already available)
Vertical short videos optimized social media
Parental controls available
Strict safety: CSAM, deepfakes blocked
Disney deal: 200+ licensed characters ($1B)
Use Cases
Social media: TikTok, Reels, YouTube Shorts
Concept reels and stylized shorts
Product teasers and brand campaigns
Pre-visualization for film/video
Storyboarding and rapid concepting
Educational explainers
Tutorial content and lesson visuals
Short films with dialogue
Creative character-driven animations
Marketing explainer videos
Client presentations
Pitch decks with visual prototypes
Mood boards and concept testing
Viral creative content
Community video sharing
Iterative animation workflows
Visual storytelling
Brand snippets
Dialogue-led bits
Cameo-based content creation
User Reviews
Related AIs
Midjourney
Midjourney Inc.
Leading AI image generator in artistic quality that transforms text prompts into stunning visual artwork, with V7 model, V1 video generation, and 21M+ user community.
Stable Diffusion
Stability AI
Open-source AI image generation model from Stability AI. Includes SD 3.5 with 8.1B parameters, runnable locally on consumer hardware, with over 10,000 fine-tuned models and free license for commercial use.
Runway
Runway AI Inc.
Leading AI video generation platform for film and creatives. Gen-4.5 (#1 Video Arena), partnerships with Lionsgate/IMAX, 300K+ customers and $3B+ valuation.