Secrets AI Video Generator: How It Works, Quality, and Cost
Video generation from AI companion images is something most platforms in this category simply do not offer. Character.AI has no video feature. CrushOn AI has no video feature. Janitor AI has no video feature. Secrets AI built it in, and it is a meaningful differentiator — not because the technology is flawless, but because the ability to animate your companion from a static image has no equivalent at a comparable price point from mainstream competitors. This guide covers the complete mechanics: how video generation works, what it costs in Moments, how the quality holds up, and when it is and is not worth spending your Moments on.
What Is the Secrets AI Video Generator?
The video generator is a feature that converts AI companion images into short animated video clips. You provide a source image (generated by the platform) and a text prompt describing the desired movement or action, and the system processes those inputs into a video clip.
This capability positions Secrets AI differently from the majority of the AI companion market. Most platforms in this segment — including well-funded competitors like Candy AI (limited video), CrushOn AI (no video), and Janitor AI (no video) — do not offer this feature at all. In the broader landscape of AI-generated content, video generation typically requires dedicated tools like Stable Diffusion video pipelines or separate subscription services. Secrets AI packages this into the companion experience.
The feature is available on the Lite tier and above — it is not accessible on the free plan. This is an important distinction for users evaluating the platform from the free tier, where no video generation capability exists.
For context on all platform capabilities beyond video, the full feature overview and complete review cover the complete picture.
How Video Generation Works
The video creation process has four steps:
- Generate or select an existing companion image — the source image forms the visual basis for the video. Higher-quality source images produce better video output; use the Premium generation model for best results.
- Add a text prompt — describe the desired movement, action, expression, or scenario. Keep the prompt specific enough to direct output but not so complex that it conflicts internally ("slow turn toward camera, soft smile" is more reliable than "turn while jumping and laughing and waving").
- Wait for processing — approximately 2 minutes for the AI to generate the clip. This is the standard processing window; complex prompts or peak-hour usage may extend it slightly.
- View and save the completed clip — review the output and save it to your account if satisfied.
Videos are short clips by design. The Lite tier produces 3-second clips; longer clips are available on higher tiers with Premium and Ultimate producing the best quality outputs. The video reflects the character's appearance from the source image and interprets the prompt as a motion or expression instruction.
Context from your current conversation scenario is incorporated where relevant — the AI attempts to align the visual output with the relationship context established in your chat.
Video Quality Assessment
Independent reviewers rate Secrets AI video quality at 4.1/5 — described as looking "good and moving smoothly most of the time." Breaking this down across what reviewers consistently report:
Strengths:
- Realistic character movement in the majority of outputs
- Natural facial expressions in standard prompts
- Visual consistency with the source image (character identity maintained)
- Smooth transitions without obvious artifacting in most clips
Weaknesses:
- Quality varies with prompt complexity — highly specific or unusual motion instructions produce less reliable results
- Occasional visual inconsistencies around hands and extremities (a common artifact in diffusion-model-based video)
- Quality ceiling is lower than dedicated AI video generation platforms (SweetDream AI, Xotic AI with 4K 15-second clips)
A score of 4.1/5 is genuinely respectable for in-platform video generation, particularly at this price point. The benchmark comparison is not against professional video tools; it is against the alternative of having no video capability at all — which is the reality at most competitor platforms.
Quality improves meaningfully when using the Premium generation model on Premium or Ultimate tiers. The standard model produces workable output; the premium model produces noticeably sharper, more fluid results.
How Much Do Videos Cost in Moments?
Video generation is the most Moments-intensive feature on the platform. The cost structure:
| Clip Type | Cost | Context |
|---|---|---|
| Short 3-second clip | ~50 Moments | Available on Lite+ |
| Standard/longer clip | ~600 Moments | Available on all paid tiers |
The practical budget impact by subscription tier:
| Tier | Monthly Moments | Short clips possible | Long clips possible |
|---|---|---|---|
| Lite | 1,000 | ~20 | ~1-2 |
| Plus | 3,000 | ~60 | ~5 |
| Premium | 8,000 | ~160 | ~13 |
| Ultimate | 15,000 | ~300 | ~25 |
Pro Tip: The jump from a short clip (~50 Moments) to a long clip (~600 Moments) is a 12x cost multiplier for what is often a modest quality or length improvement. For users on Plus (3,000 Moments/month), generating 5 long clips exhausts the entire monthly allocation. Start with short clips to test quality and prompt effectiveness before committing to long-clip generation.
For users who want to generate video heavily throughout the month, Ultimate ($39.99/mo) is the only tier where 10+ long clips per month is sustainable without frequent Moments top-up purchases.
If your Moments run short mid-month, top-up bundles start at 1,980 Moments for $5.99. Premium and Ultimate members receive bonus percentages (10% and 15% respectively) on top-up purchases.
Video vs Images vs Voice — Cost Comparison
Understanding the relative Moments cost of video versus other media features helps with budget planning:
| Feature | Cost (Moments) | Output |
|---|---|---|
| Text message | 1-2 | Single response |
| Image generation | 25-50 | Single static image |
| Short video (3s) | ~50 | Brief motion clip |
| Full video clip | ~600 | Longer animated clip |
| Voice call | 100/minute | Real-time audio |
| Manual memory save | 10 | Flagged memory |
The trade-off comparison is direct: for the same 600 Moments as one long video, you could generate 12-24 images, or have 6 minutes of voice call, or send 300-600 text messages.
This is not a reason to avoid video — it is a reason to be strategic about when video adds value over images. For capturing a specific expression or a simple action from an existing image, a short clip at 50 Moments is often the right tool. For longer narrative sequences or showcase content, long clips justify the cost. For users primarily interested in visual companion content without motion, images deliver more content per Moment.
Tips for Better Video Results
A Closer Look at what produces better outputs versus worse ones, based on how diffusion-based video models respond to input:
- Use high-quality source images — video quality is bounded by the source image quality. Generate images with the Premium model before using them as video source material.
- Start with short clips — test your prompt and character combination at 50 Moments before committing to a 600-Moment long clip. Most prompts can be validated at the short clip level.
- Keep prompts motion-focused and specific — "slow turn toward camera, soft smile, hair moving gently" works better than abstract descriptions or multiple simultaneous actions.
- Avoid overly complex multi-action prompts — the model handles one or two sequential actions more reliably than three or more simultaneous ones.
- Use the Premium generation model — available on Premium and Ultimate tiers, it produces measurably better output than the standard model.
- Generate images first, then convert the best to video — rather than video-generating from the auto-generated character images, run a dedicated image generation session and select the best output as your video source.
Who Should Use the Video Generator?
Worth the Moments investment if:
- Visual companion content is a primary reason you're using the platform
- You want something that no mainstream competitor at this price point can offer
- You plan to save or share companion content beyond screen-based viewing
- You're on Premium or Ultimate where the Moments budget makes regular video viable
Not the right focus if:
- You're on a tight Moments budget (Plus or Lite) and text/image quality matters more to you
- You're primarily a text-based user with occasional image interest
- Budget predictability is a concern — video's 600 Moments/clip cost is easy to underestimate
Best tier for video: Premium ($19.99/mo) for moderate video use (5-10 long clips/month within the 8,000 Moments allocation). Ultimate ($39.99/mo) for heavy video creators who want 15+ long clips monthly without running short.
Competitors with Video Generation
What makes the video generator genuinely notable is the competitive context:
- Character.AI: No video generation
- CrushOn AI: No video generation
- Janitor AI: No video generation
- Candy AI: Limited video (scope not publicly documented in detail)
- GirlfriendGPT: No video generation
- PocketGirl AI: No video generation
- Kalon AI: No video generation
Among platforms with comparable functionality, SweetDream AI and Xotic AI (which offers 4K 15-second clips) are the closest competitors. Neither matches Secrets AI's integration of video into the companion conversation context at the same price point.
The full Secrets AI review places this in context with the platform's overall feature stack. For how video access varies by subscription tier, the pricing breakdown and free vs premium comparison cover the exact tier requirements.
FAQ
Video length depends on your subscription tier. On the Lite plan, clips are approximately 3 seconds. On Plus, Premium, and Ultimate, longer clips are available — though the maximum length is not explicitly documented in public-facing materials. The cost scales with length: short 3-second clips cost approximately 50 Moments, while longer clips cost up to 600 Moments per clip. Processing time for any clip is approximately 2 minutes regardless of length.
No. Video generation requires a minimum of the Lite plan ($5.99/month). The free plan provides 200 starting Moments but restricts those Moments to text conversation only — no image or video generation is available at the free tier. If you want to test video generation before committing to a subscription, the Lite plan at $5.99 is the minimum investment required.
The number of videos depends on your tier's monthly Moments allocation and whether you're generating short or long clips. On the Plus tier (3,000 Moments/month): approximately 5 long clips or 60 short clips per month if Moments are used exclusively on video. On Premium (8,000 Moments): approximately 13 long clips or 160 short clips. On Ultimate (15,000 Moments): approximately 25 long clips or 300 short clips. In practice, most users split Moments across video, images, and text, reducing video volume from these maximums.
Video quality is rated 4.1/5 by independent reviewers. The outputs are described as looking "good and moving smoothly most of the time" with realistic character movement and natural facial expressions in the majority of cases. Quality varies with prompt complexity — simpler, specific motion prompts produce more reliable results than complex multi-action descriptions. The Premium generation model (available on Premium and Ultimate tiers) produces noticeably better output than the standard model. Hands and extremities occasionally show quality variations, a common artifact in diffusion-based video generation.