Affiny Roleplay Scene Video Generation — How It Works, Pricing & Tips
Affiny’s in-chat roleplay scene video generation is now live. It turns a generated still scene image into a short animated video that keeps your companion’s portrait-anchored identity — so the character you are talking to stays visually consistent even in motion.
There are two modes: Tease for quick silent clips, and Desire for longer clips with optional audio. This guide explains how the feature works, what you need, how much each mode costs, and how to get the best results.
Quick Answer
- What it does: Animates a generated in-chat roleplay scene image into a short video.
- Tease (silent): Lite 480p at 10 coins/sec; 720p at 20 coins/sec. Duration: 5 or 8 seconds only.
- Desire (optional audio): Standard 720p at 30 coins/sec; Full HD 1080p at 50 coins/sec. Duration: 2–15 seconds. Audio is a toggle at no extra coin cost.
- Requirements: Signed in, completed profile, companion with a portrait, and a generated scene image.
- No subscription required: Video generation runs on the same coin wallet as chat, voice, and images.
How It Works
The video feature is built on top of Affiny’s existing in-chat scene image generation. Instead of generating a standalone video from text, it takes a scene image you have already generated and animates it into a short clip.
The process is:
- Start or open a chat with a companion who has a portrait.
- Tap the camera icon in the chat composer to generate a scene image (15 coins). Choose Auto to use recent conversation context, or Custom to describe the scene yourself (up to 1,000 characters).
- Once the scene image is generated, select it and choose the video option.
- Pick a mode (Tease or Desire), quality, and duration.
- Affiny animates the scene using the companion portrait as the identity reference, so the character looks the same in the video as they do in their portrait and scene images.
- The finished video is posted in the chat thread with an in-character caption, just like a scene image.
Because the video starts from a still scene frame, the motion, pose, and framing are all grounded in a specific moment from your conversation rather than a generic clip.
Pricing
Tease mode (silent)
| Quality | Resolution | Cost | Duration |
|---|---|---|---|
| Lite | 480p | 10 coins per second | 5 or 8 seconds only |
| 720p | 720p | 20 coins per second | 5 or 8 seconds only |
Tease clips have no audio. A 5-second Lite clip costs 50 coins; an 8-second Lite clip costs 80 coins. A 5-second Tease 720p clip costs 100 coins; an 8-second clip costs 160 coins.
Desire mode (optional audio)
| Quality | Resolution | Cost | Duration |
|---|---|---|---|
| Standard | 720p | 30 coins per second | 2–15 seconds |
| Full HD | 1080p | 50 coins per second | 2–15 seconds |
Desire clips can include optional audio at no extra coin cost. A single Desire generation costs between 60 and 450 coins for Standard (2–15 sec), or 100 and 750 coins for Full HD.
Coins are refunded if the generation fails after deduction. If the video does not render successfully, you get the spent coins back.
What You Need Before You Start
You must be signed in and have completed your profile. This is the same account-completion gate that applies to chat, voice, and image generation.
Your companion needs a portrait. The portrait is the visual identity anchor used across scene images and videos, so the character stays recognizable.
You also need a generated scene image. Video is not generated from the chat transcript directly — it is animated from a still scene image. If you have not generated a scene image yet, tap the camera icon and create one first (15 coins).
Tips for the Best Results
Generate the scene image first, then animate it. The video inherits the pose, framing, and setting from the still image. A clearer still image leads to a cleaner video.
Use Custom mode for specific scenes. If you want a particular action, camera angle, or mood, describe it in up to 1,000 characters. The more specific the scene description, the more directed the animation will be.
Start with Tease Lite to test cheaply. A 5-second Tease Lite clip at 480p (50 coins) is the lowest-cost way to check how a scene animates before committing to Desire or longer clips.
Keep motion expectations realistic. Short clips work best for subtle motion: breathing, eye contact, hair movement, or a small gesture. Complex action over 15 seconds is harder to render consistently.
Use the companion portrait as reference. The system uses the portrait to keep the character’s face and body consistent. If the portrait is blurry or low quality, consistency in the video will suffer.
Reply to an existing scene to continue the moment. The scene library lets you pick any prior scene image and generate a new scene or video from it. This is useful for creating a sequence: a still introduction, a follow-up angle, then a short animated clip.
How It Compares to Competitors
Most AI companion platforms do not offer in-chat video generation at all. The ones that do typically handle it differently:
- Candy AI and OurDream AI have video generation, but on most tiers it is bundled into subscription plans or higher-paid tiers rather than offered as a simple coin-per-second purchase.
- Character AI, SpicyChat, Replika, and CrushOn AI do not currently offer in-chat roleplay scene video generation as a live feature.
- Affiny’s difference: video is pay-per-use, not paywalled behind a subscription. You buy coins, generate a clip when you want one, and pay only for the seconds you create. Two modes (Tease and Desire) let you choose quick silent clips or longer clips with optional audio. The same wallet covers chat, voice, portraits, scene images, and video.
That also means the feature is available immediately after signup. Your free 200 coins are enough for ~20 seconds of Tease Lite (480p) video, a short Desire Standard clip, or several scene images once your profile is complete.
FAQ
What is roleplay scene video generation on Affiny?
It is a live feature that animates a generated in-chat roleplay scene image into a short video. The video keeps the companion’s portrait as the visual identity reference, so the character stays consistent across portraits, scene images, and videos. Two modes are available: Tease (silent) and Desire (optional audio).
How much does video generation cost?
Tease (silent): Lite 480p at 10 coins/sec, or 720p at 20 coins/sec — durations fixed at 5 or 8 seconds. Desire (optional audio at no extra coin cost): Standard 720p at 30 coins/sec, or Full HD 1080p at 50 coins/sec — duration 2–15 seconds.
What do I need to generate a video?
You need to be signed in with a completed profile, have a companion with a portrait, and have generated a scene image to animate. The video is created from the scene image, not directly from the chat text.
Can I generate video without a subscription?
Yes. Video generation is coin-based, not subscription-based. You can use it with free signup coins, referral coins, or purchased coins.
What happens if video generation fails?
Coins are refunded if generation fails after deduction. You are not charged for a failed render.
How long can the videos be?
Tease clips are 5 or 8 seconds only. Desire clips are between 2 and 15 seconds. You choose the duration per clip within those limits.
Can I make a video from any scene image?
Yes. Any generated scene image in your conversation can be used as the source frame. You can also reply to an existing scene to generate a new video that continues the moment.
Is the video NSFW-capable?
Adult scene visuals are supported for fictional adults, the same as scene images. There is no separate video subscription or adult-content tier.
Does the video include sound?
Tease clips are silent. Desire clips support optional audio at no extra coin cost.
Which competitors offer in-chat video generation?
Candy AI and OurDream AI offer video generation, but it is typically tied to subscription plans. Most other AI companion platforms do not currently offer this feature. Affiny’s coin-based Tease and Desire modes are the most accessible path for users who want to pay only when they generate.
Try It Now
If you have already generated a scene image in Affiny, the video option is available now. If not, start a chat, generate a scene image from the camera icon, and animate it in seconds.