Secrets AI Video Generator: How It Works, Quality, and Cost
Video generation from AI companion images is Secrets AI's clearest competitive advantage. Character.AI, CrushOn AI, Janitor AI, and most other AI companion platforms do not offer this feature. Secrets AI does — and it works better than the skeptical expectation going in.
That said, video is also the most Moments-intensive feature on the platform. A single long clip can cost 600 Moments — the same as 12–24 images, or 6 minutes of voice calls. Whether video generation is worth your Moments budget depends on how you use the platform and which tier you're on.
This guide covers exactly how the feature works, what quality to realistically expect, and how to plan your Moments spending around it.
⚠️ Video generation on Secrets AI produces NSFW-capable content. This guide is intended for adults 18 and older.
What Makes This Feature Unusual
Most AI companion platforms are built around static interaction: text, images, voice. Motion-based content — video clips of the AI companion — is technically demanding and computationally expensive to generate, which is why so few platforms offer it.
Secrets AI's video generator converts companion images into short motion clips using AI-powered deep learning. The underlying approach is conceptually similar to how Stable Diffusion (KG: /g/11tcd8vgn9) and related diffusion models approach image generation, but extended into temporal sequences to produce motion. The result is character-consistent video that reflects the companion's appearance and the scenario context established in conversation.
The feature is available on Lite tier and above — not on the free plan. Free accounts have no access to video generation regardless of Moments balance.
For context on where video sits relative to the platform's other features, see the all features breakdown.
Step-by-Step: Generating a Video
The process is straightforward:
- Generate or select a companion image. Video generation starts from an existing still image of your companion. Higher-quality source images produce better video output.
- Open the video generation option. This is accessible from within the chat or from the image itself.
- Write a prompt. Describe the action, movement, or expression you want. Specific prompts ("walking slowly toward the camera, smiling") produce better results than vague ones ("move around").
- Submit and wait. Processing takes approximately 2 minutes per clip.
- Review and save. The completed clip can be viewed and saved.
The AI uses both the source image and the text prompt to produce character-consistent motion. The companion's appearance, clothing, and scene context carry over from the source image into the video.
Key tip: Start with shorter clips to test prompt quality before committing Moments to longer generations. A 3-second short clip at ~50 Moments is a low-cost way to verify the prompt direction before generating a longer clip at up to 600 Moments.
Quality Assessment
Video quality is rated 4.1/5 by reviewers at aigirlfriendscout — the second-highest rating among Secrets AI's features after voice (4.3/5).
What that rating reflects in practice:
- Natural movement and expressions in the majority of outputs
- Character appearance is consistent with the source image
- Smooth motion rather than jerky frame transitions
- Realistic facial expressions for straightforward prompts
Quality varies based on:
- Prompt complexity — simpler movement prompts produce more reliable results
- Source image quality — better input images produce better video output
- Generation model tier — Premium generation model produces higher-quality output than the base model
Occasional quality variations occur, particularly with complex prompts or unusual movement requests. The platform is using AI art (KG: /g/11fl0y17rl) generation techniques that still have limitations around fine-grained motion control.
The 4.1/5 rating is accurate for what the feature currently delivers: good enough to be genuinely useful and impressive compared to the field of no-video competitors, not quite at the level of dedicated video AI tools.
Moments Cost Structure
Understanding the cost range is essential because the spread is enormous — from 50 Moments to 600 Moments per clip:
| Video Type | Moments Cost | Equivalent Actions |
|---|---|---|
| Short clip (3 seconds) | ~50 Moments | Same as 1–2 images |
| Standard/longer clip | ~600 Moments | Same as 12–24 images |
This range creates very different economics depending on which clip length you're generating. A heavy video user on the Plus plan (3,000 Moments) can make approximately:
- 60 short clips (at 50 Moments each), OR
- 5 long clips (at 600 Moments each), OR
- Some combination
Monthly video budget by tier:
| Tier | Monthly Moments | Short Clips (est.) | Long Clips (est.) |
|---|---|---|---|
| Lite | 1,000 | ~20 | ~2 |
| Plus | 3,000 | ~60 | ~5 |
| Premium | 8,000 | ~160 | ~13 |
| Ultimate | 15,000 | ~300 | ~25 |
For users who want regular video generation — more than a handful of clips per month — Premium ($19.99/mo) is the practical minimum. On Lite or Plus, video generation competes directly with images and voice for a tighter Moments budget.
For a complete breakdown of which tier provides the best value for your specific usage pattern, see the video access by tier comparison.
Cost Comparison Across Media Features
The same 600 Moments can buy you very different things:
| Feature | Cost | What 600 Moments Gets You |
|---|---|---|
| Text messages | 1–2 each | 300–600 messages |
| Images | 25–50 each | 12–24 images |
| Short video (3s) | ~50 each | ~12 short clips |
| Long video | ~600 | 1 long clip |
| Voice call | 100/minute | 6 minutes of audio |
The trade-off is stark. One long video clip costs as much as a full month of casual text conversation, or 12–24 companion images. Video is a premium feature in the literal sense — it should be budgeted as such rather than consumed casually.
For users who primarily want companion images rather than video motion, the same Moments budget produces significantly more output. The Moments costs guide covers all action costs in detail.
Tips for Better Video Results
Practical improvements based on platform behavior:
- Use high-quality source images. The cleaner and more detailed the source image, the better the video output. Use the Premium generation model for source images if you plan to video them.
- Write specific movement prompts. "Turn slowly to the left, smiling at the camera" works better than "move." The AI responds to concrete motion direction.
- Test with short clips first. At ~50 Moments per 3-second clip versus ~600 for a long clip, testing prompt quality on short clips is significantly more economical.
- Avoid overly complex prompts. Multiple simultaneous actions in one prompt introduce quality uncertainty. One clear action direction per generation produces more reliable results.
- Generate from your best images. Not every auto-generated character image will be the right source for video. Select the image that best represents the visual you want in motion.
Who Should Use Video Generation
Worth the Moments cost if:
- Visual companion content is a meaningful part of why you use the platform
- You generate a moderate number of images and want some to come alive
- You're on Premium or Ultimate where 8,000+ Moments means video doesn't consume your entire budget
- The novelty of personalized AI companion video has value to you specifically
Less worthwhile if:
- You're primarily a text-based user who rarely generates images
- You're on Lite or Plus and media generation already strains your monthly Moments
- You're budget-constrained and each 600-Moment long clip feels like a significant decision
Recommended tier for video generation:
- Moderate use: Premium ($19.99/mo) — 8,000 Moments supports ~13 long clips or many more short clips monthly without exhausting the budget
- Heavy use: Ultimate ($39.99/mo) — 15,000 Moments for high-volume video creation
How Secrets AI Compares on Video
The competitive landscape is thin:
| Platform | Video Generation | Notes |
|---|---|---|
| Secrets AI | Yes | Full feature, 50–600 Moments per clip |
| Character.AI | No | — |
| CrushOn AI | No | — |
| Janitor AI | No | — |
| Candy AI | Limited | More restricted capability |
| SweetDream AI | Yes | Comparable; less mainstream |
| Xotic AI | Yes | 4K, 15-second clips; niche platform |
Mainstream AI companion platforms overwhelmingly lack video generation. The two platforms with comparable capabilities (SweetDream AI, Xotic AI) are niche products with smaller user bases and less established review data.
For users who specifically want video generation as part of their AI companion experience, Secrets AI is the mainstream platform that delivers it. This is documented in the full review as the platform's strongest competitive differentiator.
Try Secrets AI's Video Generator →
FAQ
Short clips are approximately 3 seconds. Longer clips extend beyond that, with cost increasing accordingly — short clips cost around 50 Moments while longer clips can reach up to 600 Moments. The Lite plan is limited to 3-second clips; Plus and higher tiers access the full range of clip lengths.
No. Video generation is not available on the free tier regardless of Moments balance. The feature requires a Lite plan or higher. Free accounts start with 200 one-time Moments, but none of those can be applied to video generation — only image generation and voice on paid tiers become accessible once you upgrade.
It depends on your tier and which clip length you're generating. On Premium (8,000 Moments): approximately 13 long clips or 160 short clips per month, or a mix of both. On Plus (3,000 Moments): approximately 5 long clips or 60 short clips per month. Heavy video users on Ultimate (15,000 Moments) can generate approximately 25 long clips or 300 short clips per month.
Reviewer rating: 4.1/5. Most outputs show natural movement and character-consistent appearance. Quality is best described as "good for AI-generated companion video" — noticeably better than nothing and impressive given how few platforms offer this at all, but not at the fidelity level of professional animation or high-budget video production. Prompt complexity and source image quality significantly influence output quality.