Secrets AI Video Generator: How It Works, Quality, and Cost
Video generation from AI companion images is Secrets AI's clearest competitive advantage. Most AI girlfriend platforms — including Candy AI, CrushOn AI, Janitor AI, and Character.AI — do not offer this capability in any form. Secrets AI does, and it works. This page covers the generation process step by step, realistic quality expectations, the Moments math, and whether it is worth your budget allocation.
What Is the Secrets AI Video Generator?
The video generator converts static AI companion images into short motion clips. You start with an existing image of your companion, write a text prompt describing the movement or action you want, and the system generates a video clip. The output reflects both the visual characteristics of the source image and the motion described in your prompt.
This feature is available on Lite tier and above — the free tier does not support video generation. Secrets AI is genuinely one of very few AI companion platforms offering this; it is a meaningful differentiator in the category, not a checkbox feature.
The practical use case: you have built a character you like and generated a set of images. Video generation takes those static images and brings them into motion, creating personalized visual content that no screenshot or still image can replicate.
How Video Generation Works
The process has four steps:
Step 1 — Generate or select a source image. You need an existing companion image to work from. If you do not have one yet, generate one first (25-50 Moments). The video output quality correlates with the source image quality — use your best images as source material.
Step 2 — Enter a text prompt. Write a description of the movement or action you want. Prompts can be explicit (for NSFW clips) or non-explicit (expressions, gestures, movement in a scene). Specific prompts produce better results than vague ones.
Step 3 — Wait approximately 2 minutes. Video generation takes around 2 minutes per clip. This is not instantaneous — the AI processes the image and motion request through a deep learning pipeline before rendering the output.
Step 4 — View and save the clip. The generated video appears in your chat or content area. You can view it, save it locally, or use it as a source for further generation.
Video clips are short — 3 seconds on the Lite tier, longer formats on higher tiers. They reflect the character's appearance as established in the source image and respond to the context of your conversation scenario.
Video Quality Assessment
Reviewer scores place video quality at 4.1/5. The assessment from third-party review testing: "Videos look good and move smoothly most of the time." Specific quality characteristics:
- Character movement is fluid rather than jerky in most outputs
- Facial expressions are realistic and consistent with the source image
- Character appearance is maintained accurately across the clip
- Prompt responsiveness: specific, clear prompts produce on-target output; complex or ambiguous prompts sometimes produce inconsistent results
- Quality improves on Premium and Advanced generation models
Occasional quality variations occur depending on prompt complexity and source image characteristics. Starting with high-quality source images and clear, specific prompts produces the most consistently good results.
AI-generated video at this quality level uses deep learning techniques comparable to video diffusion models, though Secrets AI does not publish technical specifications on its pipeline. The output quality is above what most users expect from AI-generated video and below broadcast production quality — appropriate for personal companion content.
How Much Do Videos Cost in Moments?
Video is the most Moments-intensive feature on the platform. Understanding the cost before generating heavily prevents budget surprises.
| Video Type | Moments Cost |
|---|---|
| Short clip (3 seconds) | ~50 Moments |
| Standard/longer clip | ~600 Moments |
Budget impact by tier (video only):
| Tier | Monthly Moments | Short Clips (50 ea) | Long Clips (600 ea) |
|---|---|---|---|
| Lite | 1,000 | ~20 clips | ~1-2 clips |
| Plus | 3,000 | ~60 clips | ~5 clips |
| Premium | 8,000 | ~160 clips | ~13 clips |
| Ultimate | 15,000 | ~300 clips | ~25 clips |
These figures assume Moments are spent entirely on video. In practice, most users split their allocation across text, images, voice, and video — which reduces the available video budget.
Key insight for heavy video users: The Moments math strongly favors the Ultimate plan ($39.99/month) if video is your primary use case. On Plus, 5 long clips per month is a tight allocation for any user who wants regular video content. Premium provides meaningful video volume (13 long clips) while maintaining budget for images and voice.
Additional Moments can be purchased separately starting at 1,980 Moments for $5.99. Premium and Ultimate subscribers receive 10% and 15% bonus Moments on all top-up purchases respectively.
Video vs Images vs Voice — Cost Comparison
| Feature | Moments Cost | Output |
|---|---|---|
| Text message | 1-2 | Text response |
| Image generation | 25-50 | Static image |
| Short video (3s) | ~50 | Brief motion clip |
| Full video | ~600 | Longer motion clip |
| Voice call | 100/minute | Real-time audio |
For the same 600 Moments, you can generate: 1 long video OR 12-24 images OR 6 minutes of voice call. Images offer the best Moments-to-content volume; long video has the highest per-clip cost but produces media that no other format matches.
For specific tier pricing and Moments allocation details, see the Moments costs page.
Tips for Better Video Results
From practical experience generating clips across multiple quality settings:
- Use your best source images. Video generation inherits the quality limitations of the source. Blurry or low-quality images produce lower-quality clips.
- Be specific in prompts. "Natural head turn with a slight smile" produces better results than "moving." The more precisely you describe the action, the more accurately the system delivers it.
- Test with short clips first. Generate a 3-second clip (50 Moments) to validate the prompt before spending 600 Moments on a full-length clip.
- Use Premium generation model. If your tier supports it, select the Premium or Advanced generation model for noticeably improved output quality.
- Match prompt to character. Prompts that align with the character's established personality and appearance context produce more coherent outputs than out-of-character requests.
- Generate images first. Before doing any video generation, build a library of high-quality images from different poses and scenarios. This gives you better source material to work from.
Who Should Use the Video Generator?
Worth the Moments if:
- You value personalized visual content alongside conversation
- You want media that reflects your specific companion rather than generic AI video
- You enjoy sharing or saving companion content for later
- Visual media is part of what you pay for on an AI companion platform
Skip or budget carefully if:
- You are primarily a text-based user and media is secondary
- You are on a tight Moments budget (Plus or Lite)
- You are not interested in visual companion media generally
- You want to maximize conversation sessions rather than media output
Best tier for video generation:
- Casual video users: Plus ($9.99) — 3,000 Moments covers occasional use
- Regular video users: Premium ($19.99) — 8,000 Moments supports consistent generation
- Heavy video creators: Ultimate ($39.99) — 15,000 Moments for high-volume output
For a full comparison of video access across tiers, see the video access by tier breakdown.
Competitors with Video Generation
The competitive landscape for AI companion video generation is genuinely sparse:
| Platform | Video Generation | Notes |
|---|---|---|
| Secrets AI | Yes | 50-600 Moments per clip, ~2 min generation |
| Character.AI | No | No video generation |
| CrushOn AI | No | No video generation |
| Candy AI | Limited | Some video features, less developed |
| Janitor AI | No | No video generation |
| SweetDream AI | Yes | Limited comparison data available |
| Xotic AI | Yes | 4K 15-second clips (higher end) |
Secrets AI video generation is a genuine differentiator. Among mainstream AI companion platforms accessible to most users, it remains one of the very few offering this feature at all. For users who want video generation as part of their AI companion experience, the platform has few direct competitors at comparable price points.
For the broader feature picture, visit the full review or the all features page.
How long are Secrets AI videos?
Video length depends on your subscription tier. The Lite plan supports short 3-second clips (approximately 50 Moments each). Plus and above access longer formats that can run up to several hundred Moments per clip for full-length videos. The exact maximum length for longer clips is not publicly specified, but the cost range (up to 600 Moments) corresponds to the higher-end output.
Can I generate video on the free plan?
No. Video generation requires at least the Lite tier ($5.99/month) and sufficient Moments. The free tier provides 200 one-time starting Moments and does not support video generation. To access video, you must upgrade to a paid plan and have enough Moments in your balance to cover the clip cost.
How many videos can I make per month?
It depends on your plan's Moments allocation and how long the clips are. On Plus (3,000 Moments), you can generate approximately 60 short clips (50 Moments each) or 5 long clips (600 Moments each) if you spend your entire Moments budget on video. In practice, most users split Moments across images, voice, and video — reducing the available video count. Premium (8,000 Moments) supports approximately 13 long clips with Moments remaining for other content.
Are the videos realistic?
Quality is rated 4.1/5 by third-party reviewers. Movement is smooth and facial expressions are realistic in most outputs. The quality is above what most users expect from AI-generated video — better than many tools in the AI art category — but it is not broadcast-quality footage. Best results come from high-quality source images and specific, clear prompts. Occasional quality variations occur with complex prompts or when source images have quality limitations.