What the official pages document
xMode is a hosted image generation platform whose official material documents image-to-video, image generation, face swapping, and image editing. Related profiles: SoulGen and Runway. Modalities, memory, and policies vary by service.
Documented workflow
The verified feature surface covers still-image generation, animation of a supplied still, targeted image editing, and face replacement. The official material describes training a reusable identity from uploaded photos, then generating scenes that hold that identity, with separate identities for faces, outfits, locations and poses combined into one scene. It states 2K output and batch generation.
Reference images, identity, and consent
Source-led editing applies only to images the user is entitled to process. Reference images, face replacement, and clothing transformation support legitimate synthetic-character work, and the same mechanisms can produce non-consensual sexual deepfakes. This profile records the capability and gives no method for impersonation or evasion.
Named on the official site
Section names xMode uses for its own features. These are reported as the site presents them, not verified as working capabilities:
- The Identity Engine
- AI Model Training
- Style Mixing
- Flexible Prompt Packs
Access and content policy
The official pricing page documents prepaid credits at a stated price per credit, with one credit described as one image generation and two model training packages priced in credits. It states that there are no subscriptions. The official material states end-to-end encryption and automatic deletion of source images after training. The official presentation frames the service for adult or NSFW use. Consent, identity, and prohibited-content constraints still apply, and the current official rules remain the controlling reference. The site publishes a record-keeping compliance statement and states that generated content is fictional and does not depict real individuals.
Not documented
Not covered by the 4 official sources behind this record, and left as unknown:
- Acceptable-use terms and enforcement
- Generation model and video clip length
- How identity training data is verified for consent