Secrets AI Video Generator: How It Works, Quality, and Cost
Ask any AI companion platform user what feature they wish their platform had, and video comes up repeatedly. The ability to see your AI companion move — not just as a static image, but in actual motion — is the natural next step from image generation. Most platforms do not offer it. Secrets AI does, and this is the feature that most clearly separates it from the competition.
That separation comes with real Moments costs and specific quality expectations. This guide covers how the video generator works, what the output actually looks like, how to budget for it, and who should make it a central part of their experience.
What Makes the Video Generator Distinctive
Secrets AI's video generation converts existing AI companion images into short video clips using a text prompt. The core capability — turning static character images into motion clips — is genuinely rare in the AI companion market.
The competitive reality:
- Character.AI: No video generation
- CrushOn AI: No video generation
- Janitor AI: No video generation
- Candy AI: Limited video capability
- Replika: No video generation
Video generation is available on the Lite plan and above — it is not part of the free tier. This deliberate gating makes video a paid-tier feature, and the Moments costs associated with it mean it also requires ongoing budget management.
Among the few platforms that offer comparable features are SweetDream AI and Xotic AI (which offers 4K 15-second clips at a higher price point). Secrets AI sits in the middle ground: better than most, not the absolute top for video quality, but well-integrated into the companion ecosystem.
How to Generate a Video: Step by Step
The process is four steps and takes approximately two minutes from start to finish:
- Generate or select an existing companion image as your base visual. Higher quality source images produce better video output — this is worth keeping in mind when you generate your initial character images.
- Add a text prompt describing the movement or action you want. Specific prompts work better than vague ones. "Walking toward the camera with a slight smile" produces more predictable results than "move naturally."
- Wait for processing. Generation takes approximately two minutes. The platform uses AI-based deep learning models to animate the still image based on your prompt. You do not need to stay on the page during processing — the video will be ready when you return.
- View and save your completed clip. The video reflects your companion's current appearance, the visual context of the conversation scenario, and the motion described in your prompt.
Videos are short clips rather than extended sequences. Length varies by tier and quality selection, from 3-second clips on Lite up to longer outputs on higher plans.
What the Quality Actually Looks Like
Video quality is rated 4.1 out of 5 by independent reviewers — described in most assessments as "looking good and moving smoothly most of the time." What this means in practice:
What works well:
- Character facial expressions and subtle movements render naturally
- The companion's distinctive appearance (hair color, style, features) is preserved from the source image
- Movement generally matches the prompt direction without obvious visual artifacts
- Output quality improves visibly with the Advanced generation model available on higher tiers
Where quality varies:
- Complex prompts requiring multiple simultaneous actions can produce inconsistent outputs
- Background elements are less stable than character rendering in motion
- Prompt complexity affects quality more than length — a specific, single-action prompt outperforms an ambitious multi-action one
The 4.1/5 rating is honest: this is not cinematic quality, and it is not the most technically advanced AI video generation available. But it is well-integrated, accessible, and functional — which is more than the same users' alternatives can say.
Moments Cost: The Real Budget Math
This is where most users need concrete numbers before committing to the feature.
| Video Type | Moments Cost |
|---|---|
| Short 3-second clip | ~50 Moments |
| Standard / longer clip | ~600 Moments |
| Image (for comparison) | 25–50 Moments |
| Voice call (for comparison) | 100 Moments/minute |
The 50–600 Moments range is wide, and it reflects meaningfully different clip lengths and quality levels. Short clips on Lite are at the lower end; full-length premium quality clips on higher tiers run toward 600.
What your monthly Moments allocation actually buys for video:
| Plan | Monthly Moments | Short Clips (~50) | Full Clips (~600) |
|---|---|---|---|
| Lite | 1,000 | ~20 | ~1-2 |
| Plus | 3,000 | ~60 | ~5 |
| Premium | 8,000 | ~160 | ~13 |
| Ultimate | 15,000 | ~300 | ~25 |
The key insight: Full-length video clips are Moments-intensive. A Premium user who generates one full clip per day will consume approximately 18,000 Moments in a month — beyond even the Ultimate allocation. Heavy daily video users need to budget additional Moments top-ups.
For the same 600 Moments as one full video clip:
- 12–24 image generations, or
- 6 minutes of voice calls, or
- 300–600 text messages
Video Versus Images Versus Voice: Where to Spend
| Feature | Cost | Best Used For |
|---|---|---|
| Text message | 1–2 Moments | Core conversation (always available) |
| Image | 25–50 Moments | Visual content, character customization |
| Short video (3s) | ~50 Moments | Quick motion clips, same cost as 1 image |
| Full video | ~600 Moments | Extended motion, high visual impact |
| Voice call | 100/min | Audio immersion, real-time feel |
For casual media users: images at 25–50 Moments each are the best value entry point. They are less Moments-intensive than full video and still provide the visual companion experience.
For users who specifically want the video feature: start with short clips to test quality before committing to full-length generation. Short clips cost approximately the same as one image — testing quality at that rate makes sense before spending 600 Moments on a longer output.
Tips for Better Video Results
These approaches consistently produce higher-quality outputs:
- Use high-quality source images. The video is generated from an image, so image quality directly limits video quality. Generate images using the Advanced model before converting to video.
- Write specific single-action prompts. "Slowly turning her head and smiling" outperforms "looking around and moving naturally." One clear action per prompt.
- Start with short clips. Test your prompt with a 3-second clip before investing in a full-length generation. This costs ~50 Moments versus up to 600 for a longer clip.
- Use the Premium generation model. Available on Premium and Ultimate plans, it produces visibly better texture and motion fidelity.
- Save your best images. Build a library of high-quality base images specifically for video generation. Having 10–15 strong source images means you always have good inputs available.
Who Should Prioritize the Video Generator?
Worth building around if you:
- Value visual companion content alongside text conversation
- Enjoy saving and collecting media from your AI companion
- Want a platform experience that feels visually immersive, not just text-based
- Are on Premium or Ultimate where Moments allocation makes regular video feasible
Not worth prioritizing if you:
- Are primarily a text-conversation user
- Are on a tight Moments budget (video is the most expensive feature per unit)
- Only occasionally want visual content (images give similar visual satisfaction at much lower cost)
- Are on the Lite plan — 1,000 Moments supports only 1–2 full video clips per month
Best plan for regular video: Premium at $19.99/month for moderate use (10–15 clips/month); Ultimate at $39.99/month for daily generation.
For a complete breakdown of what each plan includes and how Moments allocate across features, see the pricing guide and the review.
Access Video Generation on Secrets AI
Frequently Asked Questions
Short clips on the Lite plan are approximately 3 seconds. Higher-tier plans and the full video generation option produce longer clips, with full-length videos taking significantly more Moments (up to 600 per clip). The exact duration scales with quality settings and tier — higher tiers unlock both longer output and higher visual quality through the Advanced generation model. For regular use of longer clips, the Premium or Ultimate plan provides the Moments allocation that makes this practical.
No. Video generation is gated to the Lite plan and above. Free users receive 200 Moments as a starting credit, but video generation requires a paid subscription to access regardless of Moments balance. The minimum entry point for video access is the Lite plan at $5.99/month, which unlocks 3-second clips and 1,000 Moments monthly. See the free vs premium comparison for the full feature access breakdown by tier.
It depends on your plan and the clip length you choose. Lite (1,000 Moments) supports approximately 2 full-length clips or 20 short clips. Plus (3,000 Moments) supports approximately 5 full-length or 60 short clips. Premium (8,000 Moments) supports approximately 13 full-length or 160 short clips. Ultimate (15,000 Moments) supports approximately 25 full-length or 300 short clips. These numbers assume you use your entire Moments allocation on video — in practice, most users mix video with image and voice use.
Quality is rated 4.1 out of 5 by independent reviewers. Videos render character appearances accurately, with natural facial expressions and movement in most outputs. Quality varies based on prompt complexity and source image quality. The Advanced generation model on Premium and Ultimate plans produces visibly better results than the base model. Backgrounds and complex multi-character scenes are less consistent than single-character motion. The overall quality is good but not cinematic — strong enough to serve its purpose within the companion experience, with room for improvement over time.