AI Model Usage & Data Processing

Last updated: February 2, 2026

1. Third-Party AI Models

VisionFrame integrates with industry-leading AI video generation models to provide you with high-quality video creation capabilities.

1.1 Supported Models

  • Sora 2 by OpenAI - Text-to-video and image-to-video generation
  • Veo 3.1 Fast by Google DeepMind - Short video clip generation
  • Wan 2.6 by Alibaba Tongyi - Multi-aspect ratio video generation
  • Seedance 1.5 Pro by ByteDance - Audio-visual joint generation

1.2 Model Access: We access these models through their official APIs. Video generation requests are sent to the respective model providers' servers, and results are returned to our platform.

2. Data Processing & Privacy

2.1 Data Transmission: Your text prompts, images, and parameters are securely transmitted to AI model providers via encrypted API connections (HTTPS/TLS).

2.2 Data Storage: All uploaded content (text, images) and generated videos are stored securely on our servers. We do not share your data with third parties except as necessary for AI model processing.

2.3 Model Training: We do NOT use your uploaded content or generated videos to train or fine-tune AI models. Your data is used solely for generating your requested videos.

2.4 Data Retention: Your content is stored for the duration of your account plus 30 days after deletion, except where required by law for transaction records.

3. Output Quality & Limitations

3.1 Quality Variations: AI-generated video quality may vary based on prompt complexity, reference image quality, model capabilities, and current system load.

3.2 Generation Time: Video generation typically takes 3-5 minutes, but may vary based on video duration, resolution, queue position, and model provider response time.

3.3 Imperfections: AI models may occasionally produce videos with minor visual artifacts, inconsistencies, or deviations from your prompt. We recommend reviewing and, if necessary, regenerating videos before commercial use.

4. Commercial Use & Responsibility

4.1 Content Ownership: You own the videos you generate, subject to these Terms and your payment of applicable fees. However, you are solely responsible for ensuring your use complies with applicable laws and regulations.

4.2 Rights & Permissions: You must ensure you have the necessary rights and permissions for any materials you upload (images, text prompts). This includes obtaining proper consent for using likenesses, copyrighted material, or proprietary content.

4.3 Commercial Use: Commercial use rights vary by subscription plan. Please check your plan details for specific commercial use permissions.

4.4 Review Before Use: We strongly recommend reviewing all generated content for accuracy, quality, and compliance before using it for commercial or public purposes.

5. Model Provider Terms

Each AI model provider has its own terms of service and acceptable use policy. By using VisionFrame, you agree to comply with these third-party terms:

6. Technical Implementation

6.1 API Integration: We use RESTful APIs to communicate with AI model providers. All requests are authenticated and encrypted.

6.2 Queue Management: Video generation requests are queued and processed based on availability. Higher-tier plans may receive priority queue positioning.

6.3 Webhook Callbacks: AI providers notify us when video generation is complete via secure webhook callbacks. We then download and store your videos on our secure servers.

6.4 Error Handling: If generation fails due to model provider issues, policy violations, or technical errors, we provide error messages and credit refunds where applicable.

7. Compliance & Safety

7.1 Content Moderation: We implement content filtering and moderation to prevent generation of prohibited content (violence, hate speech, adult content, etc.).

7.2 Abuse Prevention: We monitor for abuse, fraud, and policy violations. Accounts violating terms may be suspended or terminated.

7.3 Rate Limiting: To ensure fair usage and system stability, we implement rate limits on video generation requests.

7.4 Audit Trail: We maintain logs of all API calls and video generation requests for security, fraud prevention, and compliance purposes.

8. Contact & Support

If you have questions about our AI model usage, data processing practices, or technical implementation, please contact us:

Contact Technical Support

We respond to technical inquiries within 1-2 business days.

Key Takeaways

  • We use industry-leading AI models through official APIs
  • Your data is encrypted and NEVER used for model training
  • You own generated content but are responsible for compliance
  • Review outputs before commercial use - quality may vary
  • Comply with both our Terms and AI providers' policies