Introduction
SoulGen operates in a category dominated by chat-first companion apps but leads with video generation. The homepage pushes the talking video angle hard: upload a reference face or write a text prompt, produce a 20-second HD clip with natural motion, immersive sound, and synchronized lip movement. That's a fundamentally different creative flow than typing messages back and forth with a character. For buyers whose interest starts with visuals rather than conversation, this positioning alone separates SoulGen from the bulk of adult AI companion platforms.
The "Looks Like" feature changes the creation pipeline in a concrete way. Upload a reference photo with a clear, frontal face, and the AI generates a character that mirrors those features — either photorealistic or anime style. This means the character you later animate in the video generator starts from a face you deliberately chose, not a random seed. For creators building a visual identity around a specific appearance, this bridges the gap between a static portrait and a living, speaking clip with recognizable continuity.
Video generation follows a three-step workflow: upload a reference image, describe the motion with a text prompt, and generate. Output maxes at 20 seconds in 720p or 1080p. The pricing page lists multiple video models — SoulGen 2.0, SoulOmni, and SoulOmni Flash — each with different credit costs per five-second segment. SoulOmni and SoulOmni Flash include audio by default, while SoulGen 2.0 charges extra for sound. Choosing between models directly affects both output cost and whether the finished clip has a voice.
Everything runs on credits. Text chat costs 0.2 credits per message. A standard image is one credit without the Looks Like reference, two with it. Video pricing scales by model, resolution, and audio — anywhere from 10 to 26 credits per five-second segment. Voice calls run one credit per minute. A ten-second SoulOmni video in 1080p, for example, consumes 48 credits before any editing passes. That arithmetic matters immediately for anyone planning to use the platform for more than a few test renders.
Beyond generation, SoulGen layers Smart Edit for natural-language image manipulation, Outpainting for extending scenes beyond their original borders, and Face Swap with a Face Enhance toggle for sharper facial definition. These tools function as post-processing steps — generate a character, swap in a different face, edit the setting, expand the composition. The Chat feature also ties into image generation: request photos during a conversation with a created Soulmate at two credits each, and the character responds with a generated image.
Here's the practical call for buyers: SoulGen suits people who want to produce short-form visual adult content, not writers seeking extended text roleplay. The video pipeline is the real product, and the credit costs reflect that priority. If your main interest is immersive conversation with a character, the per-message cost is low but media add-ons stack fast. If you're here to animate characters from specific reference faces, the Looks-Plus-Video combination is something most chat-first competitors simply don't offer.
Merits
- Video generation with lip-sync and audio from a reference image is rare in the adult AI space — most competitors stop at still images or text.
- The Looks Like feature gives you a direct pipeline from a chosen face to a generated character to an animated video with visual continuity.
- Multiple video models with different credit costs let you draft cheaply on SoulGen 2.0 and upgrade to SoulOmni for final renders.
- Smart Edit, Outpainting, and Face Swap function as a modular post-processing chain rather than forcing everything into a single generation pass.
Disadvantages
- Subscription price is not published on the pricing page — you must sign in before seeing the actual Pro membership cost.
- Video output caps at 20 seconds, which limits longer narrative scenes or extended character animation without stitching clips externally.
- Credit costs scale aggressively for higher-resolution audio-included video — sustained production at 1080p with sound burns through allowances quickly.
- Refund policy excludes sale items and any used credits, with only a 14-day window and no partial refunds.
Key Features
Image-to-Video & Reference-to-Video
Turn a static character image into a 20-second HD video with motion, audio, and lip-sync — the core pipeline that separates SoulGen from text-only adult AI platforms.
Looks Like Reference Upload
Anchor generated characters to a specific face photo at two credits per image, giving you visual consistency across stills and animated output.
Smart Edit with Text Prompts
Modify any generated image by describing changes in natural language — add, remove, or alter elements without regenerating the entire composition from scratch.
Face Swap with Face Enhance
Swap faces in images or videos and toggle an enhancement mode that sharpens facial clarity and definition for cleaner final output.
SoulChat with In-Conversation Image Generation
Hold text or voice conversations with a created Soulmate and request generated photos during the chat at two credits per image.
AI Outpainting
Resize an image to any ratio and let the AI automatically fill expanded borders with new backgrounds, characters, or scene elements.
Editor Score
SoulGen's real strength is its video pipeline — moving from a reference face to a living, talking HD clip in a handful of steps. That's a distinct offering in a market crowded with text-only chatbots. The modular editing tools add genuine utility for creators refining outputs over multiple passes. But the unpublished subscription pricing and aggressive credit scaling for high-quality video keep the total cost of ownership uncertain until you're already committed inside the system.
| Criterion | Score |
|---|---|
| Character-to-Video Pipeline | 16 / 20 |
| Visual Customization Depth | 16 / 20 |
| Pricing Transparency | 10 / 20 |
| Post-Generation Editing Control | 15 / 20 |
| Credit Efficiency at Scale | 12 / 20 |
| Total | 69 / 100 |
Rate Plans
SoulGen runs on credits with a Pro membership that includes 100 credits per month, 1,200 annually. The homepage promotes up to 70% off for a first purchase, but the base subscription price is not published on the pricing page — it appears behind sign-in. Credits cover images, video, chat, voice calls, and character creation. Before committing, compare your expected workflow against the per-operation credit costs listed on the pricing page, since a single ten-second HD video with audio can cost 24 to 48 credits depending on the model.
| Plan | Price | Who it suits |
|---|---|---|
| Free | Available | Limited credits for testing image generation, basic video, and chat; free user chat memories retained for a limited number of days. |
| Pro | Shown after sign-in | 100 credits per month (1,200/year), multitask creation access, and unlocked Pro features; exact price requires account login. |
Refunds are available within 14 days of purchase only if no credits have been used since the payment. Credit packs and sale items are non-refundable with no partial or pro-rata refunds. Deleting your account with an active auto-renewing subscription cancels it automatically.
Alternative Tools
| Tool | Why compare it |
|---|---|
| CrushOn AI | SoulGen is built around visual output — video, images, face swap — while CrushOn AI is a text-first roleplay platform with character cards, model switching, group chat, and shared world-building. If your priority is extended narrative roleplay with multiple characters in one conversation, CrushOn offers deeper chat architecture and community-created cards. If you want to generate and animate specific visual characters with lip-sync, SoulGen's video pipeline is the clear differentiator. |
| GirlfriendGPT | GirlfriendGPT focuses on browsing and chatting with a large catalog of adult AI personalities, with custom character creation built around personality, scenario, and opening situation. SoulGen trades that browsing-and-chat experience for a visual content production pipeline — reference faces, video generation, and layered image editing. Choose GirlfriendGPT for character-driven text interaction across many personas; choose SoulGen for producing short visual content anchored to specific face references. |
| OurDream AI | Both SoulGen and OurDream AI combine chat, images, and video around AI-generated companions. OurDream attaches all media to one persistent character with voice calls and image-to-video for a flat 100 coins per five-second standard clip. SoulGen offers similar breadth but with more granular model choices — SoulOmni, SoulOmni Flash, and SoulGen 2.0 — letting you trade quality for cheaper draft renders. The practical difference is pricing control at each resolution tier. |
Frequently Asked Questions
What video resolutions and lengths does SoulGen support?
SoulGen generates videos in 720p and 1080p, up to 20 seconds per clip, across multiple models with different credit costs.
How does the Looks Like feature work?
Upload a reference photo with a clear, frontal face. The AI generates a character that mirrors those features in photorealistic or anime style at two credits per image instead of the standard one.
Can I use SoulGen-generated content commercially?
Yes. The Terms of Service state that AI-generated videos may be used for commercial purposes, provided the content complies with applicable laws and does not infringe on third-party rights.
How long is my generated content stored?
Content not published to the community is stored for 7 days, then automatically deleted. Published content remains on the server until you delete it manually.
Is there a free tier?
Yes. Free users receive limited credits for testing image generation, basic video creation, and chat. Free user chat memories are saved for a limited number of days.
What's the refund policy?
Subscription refunds are available within 14 days if no credits have been used since the payment. Credit packs and sale items are non-refundable. No partial or pro-rata refunds are offered.