18+

Secrets AI Video Generator: How It Works, Quality, and Cost

Among mainstream AI companion platforms, video generation is genuinely scarce. Character.AI, CrushOn AI, Janitor AI, and GirlfriendGPT do not offer it. Candy AI has limited video. Secrets AI offers video generation as a core feature across its paid tiers — which makes it the most capable platform in this space for users who want their companion to move.

This page covers the mechanics, quality, cost structure, and honest assessment of when the video generator is worth using versus when Moments are better spent elsewhere.

For context on the broader platform this feature lives within, the full review covers everything.

What the Video Generator Does

The video generator turns static AI companion images into short motion clips. The starting point is an existing companion image — either one of the 4 images auto-generated when you create a character, or any subsequent image generated during chat. You add a text prompt describing the desired movement or action, submit the request, and wait approximately 2 minutes for the clip to generate.

The output is a video clip that reflects both the companion's appearance (from the source image) and the motion or expression described in the prompt. Clips are context-aware in the sense that the character's established appearance, clothing, and setting carry through from the source image into the video — you're animating an existing companion, not creating a generic video.

Availability: free users cannot generate video. Lite plan ($5.99/month) enables 3-second clips. Plus ($9.99) and above unlock longer clips and better quality output through Advanced generation models on Premium/Ultimate.

The Generation Process Step by Step

  1. Generate or select an existing companion image from your chat or character gallery
  2. Tap or click the video generation option associated with that image
  3. Enter a text prompt describing the motion, action, or expression you want
  4. Submit the request — generation begins immediately
  5. Wait approximately 2 minutes for processing
  6. View the completed clip in your chat or media gallery
  7. Save or share the clip as desired

Prompts that specify distinct, singular actions produce cleaner results than multi-action prompts. "Turn and smile toward the camera" produces better output than "turn, smile, reach forward, then look away." Specificity in what the motion is — rather than complexity in how many motions there are — correlates with quality.

Video Quality: Honest Assessment

Independent reviewers rate Secrets AI video quality at 4.1/5. The practical description from reviewer testing: videos "look good and move smoothly most of the time." Facial expressions are recognizable and appropriately matched to the prompt description. Character movement is realistic rather than mechanical in most outputs.

Factors that affect output quality:

  • Source image quality — higher-fidelity source images produce better video output; blurry or low-contrast source images carry those flaws into the motion output
  • Generation model — Advanced model (Premium and Ultimate plans) produces meaningfully better results than the standard model
  • Prompt specificity — clear, singular action descriptions outperform vague or complex multi-action prompts
  • Prompt complexity — highly specific facial expression combinations or complex scene interactions sometimes produce inconsistent results

Quality is not uniform across every generation. Approximately 1 in 5-6 clips will have a notable quality issue — a movement that looks slightly unnatural, facial rendering that loses some fidelity, or a motion that doesn't match the prompt intent precisely. The 4.1/5 rating reflects solid average quality rather than guaranteed perfection on every generation.

At current AI video generation capabilities (2026), Secrets AI's output quality sits above what free tools offer and below what dedicated AI video platforms targeting professional output produce. For companion interaction and personal media use, the quality is appropriate and consistently appreciated by reviewers.

Moments Cost by Clip Type

Video is the most Moments-intensive feature on the platform:

FeatureMoments CostComparison
Text message1-2Cheapest
Image generation25-50Moderate
Short video (3 seconds)~50Similar to 1-2 images
Full video clip~600Significant
Voice call100/minutePer minute

The 600-Moment cost for a full video clip means:

  • On Lite (1,000 Moments/month): 1-2 full videos, or up to 20 short clips — that's the entire monthly budget
  • On Plus (3,000 Moments/month): up to 5 full videos, or 60 short clips — leaving little room for images and voice
  • On Premium (8,000 Moments/month): up to 13 full videos, with budget remaining for images
  • On Ultimate (15,000 Moments/month): up to 25 full videos comfortably

For users who want regular video generation as a primary use case, the math is clear: Ultimate ($39.99) or Premium ($19.99) are the practical tiers for video. Lite and Plus can support occasional video, but frequent generation depletes the monthly allocation quickly.

The video access by tier breakdown in the free vs premium comparison covers this budget math in the context of the full feature set.

Comparing Moments Spend: Video vs Other Features

For the same 600 Moments that a single full video clip costs, you could alternatively generate:

  • 12-24 images (at 25-50 Moments each)
  • 6 minutes of voice calls (at 100 Moments/minute)
  • 300+ text messages

The trade-off is genuine: video is dramatically more Moments-intensive than any other feature. Users who primarily value visual media content should generate images to select the best candidates before converting them to video — this process maximizes video quality (better source image) while minimizing wasted Moments on video attempts from suboptimal sources.

Strategy for Getting the Most from the Video Generator

Based on the Moments economics and quality factors:

  • Generate images first, video second. The 4 auto-generated character images are free starting points. Generate additional images from the best source material before spending Moments on video conversion.
  • Start with short clips. A 3-second clip (~50 Moments) lets you test whether a particular prompt and source image combination produces good results before committing to a full 600-Moment clip.
  • Keep prompts focused. Describe one action clearly. "Laugh naturally" works better than "laugh while reaching forward and turning."
  • Use Premium or Advanced models for final clips. If you're creating video for content you plan to save or share, spend the extra Moments on the higher-quality generation model. For drafts or tests, standard model is fine.
  • Top-up strategically if needed. Additional Moments can be purchased starting at 1,980 Moments for $5.99. Premium and Ultimate subscribers receive 10-15% bonus Moments on top-ups. See the Moments costs breakdown for full top-up pricing.

Who Should Use the Video Generator

The video generator adds genuine value for certain user types and limited value for others:

Worth using if:

  • Visual media is a significant part of why you use the platform
  • You want motion content from your companion rather than static images only
  • You're willing to invest in Premium or Ultimate to have a workable Moments budget
  • You're interested in creating content to save and revisit over time

Not worth using if:

  • You're on Lite or Plus and primarily use text conversation — video will drain your entire Moments budget quickly
  • You're in the evaluation phase on free — video isn't available and isn't worth buying Moments for before you've confirmed the platform fits
  • You primarily use voice calls — 600 Moments supports 6 minutes of voice, which may provide more value than 1 video clip

How Secrets AI Video Stacks Up Against Competitors

The competitive picture for video generation among AI companion platforms:

PlatformVideo GenerationNotes
Secrets AIYes — full feature50-600 Moments per clip
Candy AILimitedLess developed than Secrets AI
Character.AINoNo video at any tier
CrushOn AINoNo video generation
Janitor AINoNo video generation
GirlfriendGPTNoNo video generation
SweetDream AIYesComparable, smaller platform
Xotic AIYes4K, 15-second clips — more specialized

Secrets AI's video generation is genuinely distinctive in the mainstream AI companion space. The only comparable options are smaller, more specialized platforms (SweetDream AI, Xotic AI). Among the major platforms, Secrets AI stands alone.

This is the strongest argument for choosing Secrets AI over alternatives for users who care about visual media: no mainstream competitor matches its video capability. For users who don't care about video, this differentiator is irrelevant — and competitors like CrushOn AI (lower price) or Candy AI (better static image quality) may serve them better. The all features page and pricing help make that comparison.

FAQ

Short clips on Lite plan are approximately 3 seconds. Longer clips, available on Plus and above, run for an extended duration with better quality output. Full clips cost up to ~600 Moments. There is no fixed maximum length documented — "full video" is the platform's category for non-short clips, with cost and duration varying by generation parameters.

No. Video generation requires at minimum a Lite subscription ($5.99/month). Free accounts do not have access to video generation. The 200 starting Moments on free accounts are better used for testing image generation or saving for manual memory actions during text conversations.

Depends on your tier and Moments allocation. On Lite (1,000 Moments): approximately 2 full clips or up to 20 three-second clips. On Plus (3,000 Moments): approximately 5 full clips or 60 short clips. On Premium (8,000 Moments): approximately 13 full clips. On Ultimate (15,000 Moments): approximately 25 full clips. These estimates assume you spend all Moments on video — in practice, mixing with images and voice produces fewer total videos per month.

Rated 4.1/5 by independent reviewers. The consensus description is that they "look good and move smoothly most of the time." Facial expressions and character movement are realistic in most outputs. Quality varies based on source image quality, prompt specificity, and generation model tier (Advanced models on Premium/Ultimate produce better results). Approximately 1 in 5-6 clips has a notable quality issue. The overall quality is appropriate for companion interaction use, not professional production.

Get Started