Video Generation on Affiny: Tease and Desire Explained

Video generation turns a still image into a short moving clip. Your companion keeps the same face from their portrait, so video matches the images you have already made. Available on Companion and Intimate plans only. For prompt patterns and motion examples, see video generation tips.

Quick answer

Video generation is a subscriber-only feature on Affiny. Pick a scene still or your companion's portrait, tap the film icon in the message box (Generate video), choose Tease (short, silent, from 5 per second) or Desire (longer, optional sound, from 30 per second), describe one clear motion, tap Generate Video, and wait for the clip in chat. The video always starts from the exact still you give it, so a sharp first frame matters more than a long motion prompt. If generation fails, your coins come back automatically.

Who can make videos

Scene videos require a Companion or Intimate subscription. Both plans include the same video features. They differ only in monthly coins and how many companions you can create.

Free users see chat, voice, and image generation as usual. When you tap Generate video in the message box or ask for a clip in chat without a subscription, Affiny shows an upgrade prompt instead of the Generate Video panel.

Subscribers also pay less per still (10 vs 15), which makes the image first, then animate workflow cheaper in practice. You can build a strong first frame on a plan, or save stills while you decide whether to subscribe.

Tease vs Desire

Both modes turn the same still into motion. They differ in length, sound, quality options, and price.

Tease suits a hair flip, a slow blink, or a small gesture. It is the cheaper way to see whether a pose animates well.

Desire suits a turn, a walk, a longer reaction, or a clip where you want her to speak. Turn sound on and put a short line (5 to 10 words) in the motion field if you want audio.

Step by step: animate a still

The most reliable path is to generate or pick a still first, then animate it.

1. Pick your first frame

Open chat and choose one of these starting points:

• A scene image you already generated in this conversation
• Any image from the chat Media gallery (images icon in the chat header, View media gallery)
• Your companion's portrait (the panel shows "Continuing from companion portrait")

The clip inherits the face, outfit, lighting, and pose from that exact frame. If the still is blurry or cropped tight on an eye, the video will too.

2. Set the video source (scene stills)

For a scene image in chat, tap Continue from this scene (the curved arrow at the bottom right of the image). A reply strip appears above the message box with the still you picked.

To pick from the gallery instead, tap Pick source from gallery (pen icon in the message box), choose a still or portrait, then continue.

For portrait-only video, skip this step and go to step 3. The Generate Video panel will show "Continuing from companion portrait".

3. Open the Generate Video panel

Tap the film icon in the message box (Generate video). The Generate Video sheet opens with Tease and Desire options, length and quality controls.

4. Choose Tease or Desire

Pick the mode that fits the motion you have in mind. Tease for short and silent. Desire when you need more time or optional sound.

5. Set length and quality

• Tease: 5 or 8 seconds. Lite costs 5 per second; Standard costs 10 per second.
• Desire: 2 to 15 seconds on the Duration slider. Standard costs 30 per second; Full HD costs 50 per second.

The Estimated cost line at the bottom of the panel shows the total before you tap Generate Video.

Match length to action. Five seconds fits one gesture. Fifteen seconds fits a slow turn, a few steps, or a short spoken line with sound on.

6. Write the motion (optional but recommended)

In the Direction box, describe one clear action in plain language. Good: "She turns her head toward the camera and smiles slowly." Weak: "She does something cool."

On Desire you can also fill Negative prompt (things to avoid, up to 500 characters) if a detail keeps showing up wrong, and tick Include sound if you want her to speak.

7. Generate and watch

Tap Generate Video and wait. The finished clip appears in chat like an image message. You can replay it from there or download it.

Step by step: ask in chat instead

You can skip the gallery and type a video request in the message box. When Affiny reads it as a video intent, the same popup as image generation appears with Image and Video buttons. Tap Video, then follow the same Tease or Desire flow.

Video requests, asking plus a video word:

• "Send me a video"
• "Show me a vid"
• "Can you send a clip?"

Picture-style asks ("send me a pic", "show me what you look like", and similar) open the same popup, so you can start with an image and animate it, or go straight to video from chat. See image generation for more natural-language examples.

What a clip costs

Video is priced per second. Multiply the rate by the length you choose. Subscribers also pay lower image rates (10 vs 15), so building a strong first frame costs less on a plan.

Image plus video combo: a subscriber might spend 10 on the still, then 25 on a 5-second Tease, for 35 total. A free user cannot animate, but can still make the still for 15 and subscribe later to animate it.

If a generation fails, coins return automatically. Only successful clips charge your wallet.

Image first, then animate

Every good clip is really two steps: make a great still, then describe one great motion. The video engine does not invent a new scene. It moves the pixels you already have.

1. Generate a scene image with the pose, outfit, and background you want. Use image generation tips for sharper results.
2. Check the still before you animate. Face clear? Hands visible? Angle suits the action?
3. Animate with one verb-led motion. "She leans forward and laughs" beats a paragraph of story.
4. Chain clips if you need a sequence. Image, then video, then a new image for the next beat.

Full prompt patterns, sound examples, and before/after fixes live in video generation tips.

If it is not working

Adult content

Adult motion between fictional adults is supported for subscribers, subject to the same safety rules as chat and images. God Mode on a subscription unlocks fully uncensored roleplay in text; video still follows platform safety limits.

Portrait vs image vs video

Your companion's portrait anchors every scene image and every video, so she stays recognisable from frame to frame.