Secrets AI Video Generator: How It Works, Quality, and Cost
Video generation on Secrets AI is the platform's most distinctive feature — and also the fastest way to exhaust your Moments allocation if you go in without a plan. At up to 600 Moments per full clip, a Premium account's monthly allocation of 8,000 Moments supports only about 13 full videos if you spend everything on them. Understanding the cost structure and quality expectations before generating your first video prevents the frustration of an empty balance after two sessions.
The Feature Most AI Companion Platforms Do Not Have
Most AI companion platforms stop at image generation. Secrets AI goes further: it generates short video clips from your companion images using a text prompt describing the movement or action you want. This is genuinely rare — Character.AI, CrushOn AI, and Janitor AI do not offer video generation. Candy AI offers limited video capability; Secrets AI's implementation is more fully developed.
Video generation is available on the Lite plan and above — it is not accessible on the free tier at all. Short three-second clips are available on Lite; longer clips require Plus or higher. The feature represents a real differentiator in this category, and it is the clearest reason to choose Secrets AI over otherwise comparable alternatives.
The underlying technology draws on deep learning image synthesis similar to approaches used in tools like Stable Diffusion for image generation, extended to produce temporal motion from a static source image. The AI generates motion sequences frame-by-frame based on your text prompt, constrained by the source image's character appearance and the established conversation scenario context.
Step-by-Step: How to Generate a Video
Generating a video requires four steps, with a wait of approximately two minutes between submission and delivery:
- Generate or select an existing companion image. The quality of the source image directly affects video quality — high-quality, clear images produce better video output. The four auto-generated character images you get at character creation are good starting points.
- Add a text prompt describing the desired movement or action. Specific, clear prompts produce more reliable results than vague or overly complex ones. "Slowly turns and smiles at camera" works better than "does a complex sequence of movements while looking at the viewer from multiple angles."
- Submit the request and wait approximately two minutes. The AI processes the request server-side during this period. Keep the browser open and active — background generation is not supported on the browser-based platform.
- View and save the completed clip. Videos can be saved to your device directly from the interface.
Clips generated on the Lite tier are limited to approximately three seconds. Higher tiers support longer clips. The video length available per tier is tied to the generation model quality, with Premium and Ultimate using the Advanced model for both better quality and longer duration.
Video Quality — What You Actually Get
External reviewers rate Secrets AI video quality at 4.1/5 — a strong score for the current state of AI-generated video, reflecting that the outputs look good most of the time while acknowledging room for improvement.
Realistic character movement and facial expressions are the consistent strengths. Characters move with natural-looking motion in most clips — gestures and expressions reflect the prompt direction without the jerky or distorted movement common in lower-quality AI video tools. The generation quality is particularly strong for simpler, more naturalistic movements.
Prompt complexity is the main variable affecting quality consistency. Simple, clear prompts reliably produce good results. Complex or highly specific prompts — particularly those involving multiple simultaneous actions or unusual camera angles — produce more variable results. Occasional artifacts and blurring appear on edges and fine details, especially in complex scenes.
Generation quality improves meaningfully when using the Premium or Advanced generation model, available on Premium and Ultimate subscription tiers. If video quality matters to your use of the platform, the model upgrade is worth factoring into tier selection.
How Much Video Generation Costs
Understanding the Moments cost structure prevents the most common disappointment: running out of allocation faster than expected.
Cost by clip type:
- Short three-second clip: approximately 50 Moments
- Standard/full-length clip: approximately 600 Moments
The range between 50 and 600 depends on clip length and quality settings. Budget planning should use 600 as the worst-case assumption for longer clips.
Monthly video budget by tier:
| Plan | Moments/Month | Short Videos (50 Moments) | Full Videos (600 Moments) |
|---|---|---|---|
| Lite | 1,000 | ~20 short clips | ~1–2 full clips |
| Plus | 3,000 | ~60 short clips | ~5 full clips |
| Premium | 8,800 (+ bonus) | ~176 short clips | ~14 full clips |
| Ultimate | 17,250 (+ bonus) | ~345 short clips | ~28 full clips |
These figures represent video-only budgets — in practice, your Moments are split across images, voice, text saves, and video. The real-world video count per month is lower if you use other media features.
Heavy video users generating more than 10–15 full clips per month should strongly consider either the Ultimate plan ($39.99/month) for the highest Moments allocation, or supplementing any tier with Moments top-up bundles, which start at 1,980 Moments for $5.99 and scale up to 118,800 Moments for $249.99.
Video vs Images vs Voice — Where Your Moments Go Furthest
| Feature | Moments Cost | What You Get |
|---|---|---|
| Text message | 1–2 | Full conversation exchange |
| Image generation | 25–50 | Single static image |
| Short video (3s) | ~50 | Brief motion clip |
| Full video | ~600 | Longer motion clip |
| Voice call | 100/minute | Real-time audio interaction |
| Manual memory save | 10 | One stored memory item |
For the same 600 Moments, you can choose:
- 1 full video clip, or
- 12–24 static images, or
- 6 minutes of voice interaction, or
- Approximately 300 text messages
Static images deliver more visual content per Moment than video. If your goal is having multiple companion visuals to browse, image generation is substantially more Moments-efficient. Video adds motion and dynamic quality that static images cannot replicate — it is a qualitatively different experience, not just a better version of images.
How to Get Better Video Results
These practices consistently improve output quality:
- Use your highest-quality source images. Generate a few images specifically for video conversion rather than using auto-generated character images — choose one with clear, well-rendered details and a natural pose that suits the motion you want.
- Keep prompts specific but not complex. "Smiles warmly and looks to the side" is better than "does a slow dramatic turn with her hair flowing while smiling and then looks back with a shy expression." One primary action per prompt produces more reliable results.
- Start with a short clip before committing to a long one. Spend 50 Moments on a three-second test clip with your prompt before spending 600 Moments on a full-length version. If the short clip shows the quality and direction you want, proceed to the longer version.
- Use the Premium generation model. If you're on Premium or Ultimate, the Advanced model produces noticeably better motion quality and fewer artifacts.
- Generate images first, then convert to video. Auto-generated images are fine, but purpose-built images for video conversion produce better output. Generate a specific pose or expression, then build the video from that.
Who Should Use the Video Generator (and Who Should Skip It)
The video generator adds real value for specific users and is not worth the Moments cost for others.
Worth using if:
- You want visual content that goes beyond static images — motion and expressions add qualitative depth to the companion experience
- You create companion content you save and revisit
- You use video as part of extended roleplay scenarios where seeing character motion enhances immersion
- You're on Premium or Ultimate with enough Moments to absorb the cost without depleting your allocation for other features
Not worth the Moments if:
- You're on a tight Moments budget and primarily use text conversation — the allocation is better spent on images or voice
- You're on the Lite plan with 1,000 Moments per month — a few full video clips consumes a significant fraction of that allocation
- Video motion doesn't add meaningfully to your use case
Best tier for regular video use: Premium ($19.99/month) for moderate use of 5–10 full clips per month. Ultimate ($39.99/month) for heavy video generation of 15+ full clips monthly.
Alternatives That Also Have Video Generation
The scarcity of video generation in this category is part of what makes the feature meaningful as a differentiator.
- Character.AI: No video generation
- CrushOn AI: No video generation
- Janitor AI: No video generation
- Candy AI: Limited video capability
- SweetDream AI: Video generation available
- Xotic AI: 4K video clips up to 15 seconds
Among widely available AI companion platforms, Secrets AI is one of very few with a developed video generation feature. SweetDream AI and Xotic AI are the closest competitors on this specific capability — Xotic AI's 4K 15-second clips are technically impressive but represent a more specialized use case. For users who consider video generation a deciding feature in platform selection, the competitive landscape strongly favors staying with Secrets AI rather than switching.
For the full Moments cost breakdown across all features, see the Moments costs and plans page. To compare how video access changes across subscription tiers, see the video access by tier breakdown.
FAQ
The most common causes are: insufficient Moments balance (full video clips cost up to 600 Moments — check your balance before generating), wrong subscription tier (video generation requires Lite or above; it is not available on the free tier), or a network interruption during the ~2-minute processing period (keep your browser open and active while generating). If none of these apply and video generation still fails, try a simpler text prompt — overly complex prompts occasionally fail to produce valid output. If the problem persists, contact support.
No. Video generation is not available on the free tier. Short three-second clips become accessible starting from the Lite plan ($5.99/month). Longer, higher-quality clips are available on Plus and above. Your 200 free starting Moments cannot be used for video — they cover images (25–50 Moments each) or voice (100 Moments/minute), but video generation requires a paid subscription regardless of Moments balance.
It depends on your subscription tier and video length. On Plus (3,000 Moments): approximately 5 full-length clips or up to 60 short clips if you spend all Moments on video. On Premium (~8,800 Moments): approximately 14 full-length clips or up to 176 short clips. On Ultimate (~17,250 Moments): approximately 28 full-length clips or up to 345 short clips. In practice, your monthly video count will be lower if you also generate images or use voice. Moments top-up bundles (from $5.99) are available if you need additional capacity mid-month.
The most reliable improvements come from four changes: use purpose-built source images rather than auto-generated defaults (generate a high-quality image specifically for video conversion), simplify your prompts (one clear action produces better results than multiple simultaneous actions), use the Premium or Advanced generation model if your tier supports it, and test with a short three-second clip before committing Moments to a full-length video. These steps collectively resolve most quality complaints.