YumeVerse · Complete User Guide

1. Platform Overview

YumeVerse is a professional-grade, uncensored AI video generation platform built for adult creators. The platform applies no content review, moderation, or manual approval to prompts, reference assets, or rendered output — enabling unrestricted creative expression, including full-scale adult-oriented content, within applicable legal boundaries.

Compared with mainstream restricted platforms, YumeVerse offers the following core differentiators across the creation pipeline:

• Zero-moderation generation — no topic blocklists, no sensitive-word filtering, no post-render human review
• No face or identity restrictions — no celebrity detection, public-figure filtering, or facial compliance gates; upload any character reference and lock appearance with high consistency weight
• Adult-oriented template ecosystem — curated 18+ presets covering common adult themes and established camera language
• Six-step structured wizard — project brief → scene → character → motion & camera → output specs → submit, balancing professional control with efficient production
• Tiered membership & on-chain billing — differentiated quotas, per-clip duration caps, and queue priority; transparent USDT (TRC20) subscription

2. Interface & Navigation

2.1 Desktop Primary Navigation

2.2 Mobile Navigation

A fixed bottom bar provides equivalent access to core modules on smaller screens.

2.3 Language Switching

Use the language toggle in the top navigation bar to switch between Chinese and English. This guide follows your selected language automatically.

3. Video Templates

3.1 Browse Templates

On Templates , each card shows a cover preview, style tags, scene/character reference thumbnails, and an expandable prompt summary.

3.2 Apply a Template

1. Click Use Template on the desired card
2. You will be taken to Video with all wizard fields pre-filled
3. Built-in scene and character references load into Steps 2 and 3
4. Override any field or upload custom references at any step

4. Six-Step Video Wizard

On Video , the left panel is the editor; the right panel is a live parameter preview. Navigate across six stages via step tabs or the footer controls.

Step 1 · Project & Overall Description

• Project name (optional) — use a series name plus segment ID, e.g. "Coastal Walk · Part 1"
• Overall video description (required) — the core creative prompt covering subject, environment, mood, and visual style

Example: "A red-haired young woman walks slowly along a beach at dusk; sea breeze moves her dress; orange-pink sunset in the distance; cinematic slow-motion feel."

Step 2 · Scene Assets

• Scene references — up to 3 images (JPG / PNG, etc.)
• Scene description — textual detail on environment, lighting, and spatial context
• Scene weight (0–1) — how closely output follows scene references; default ~0.6

Step 3 · Character Assets

• Character references — up to 4 images for appearance lock-in
• Character description — structured traits: age, hair, wardrobe, expression
• Character consistency weight (0–1) — default ~0.8
• Character count — number of figures in frame

Step 4 · Motion & Camera

• Motion / action description (required) — dynamic changes within this clip
• Camera movement — static / push / pull / orbit / track / handheld
• Shot type — close-up / medium / wide / high angle / low angle
• Transition — cut / fade / dissolve (cut is typical for single clips)

Step 5 · Output Parameters

Step 6 · Review & Submit

1. Review the JSON parameter summary; verify duration, aspect ratio, and seed
2. Click Confirm & Submit
3. Non-VIP users see a subscription prompt — proceed to VIP
4. VIP users: the job enters the processing queue; track status in History

5. Continuous & Sequential Video

Per-job duration limits apply by tier: VIP Monthly/Yearly — 15 seconds; SVIP / SVIP+ — 30 seconds. For longer deliverables or shot-to-shot continuity, use the workflows below.

Method A · Single Maximum Duration

When your target fits within the tier cap, select the maximum allowed duration in Step 5 and describe the full timeline in Steps 1 and 4 (e.g. "0–10s walk to shoreline, 10–15s pause and look back at camera").

Method B · Segmented Sequels (Recommended for 30s+)

Segment 1 · Establish the Base Clip

1. Configure Steps 1–5 in Video
2. Name the project with a segment identifier (e.g. "Series Name · Part 1")
3. Before submitting at Step 6, archive all parameters — especially seed, aspect ratio, and resolution
4. After submission, wait for completion in History and download the clip

Last-Frame Extraction · Visual Anchor

1. Open Segment 1 in VLC, CapCut, or FFmpeg
2. Export the final frame (or the clearest near-final frame) as PNG
3. This frame becomes the scene reference for the next segment

Segment 2 · Parameter-Linked Continuation

1. Return to Video (reuse the same template for quick parameter reload)
2. Update the project name suffix; add "continues from previous shot" in the overall description
3. Step 2: upload the last-frame PNG; scene weight 0.7–0.85 recommended
4. Step 3: retain Segment 1 character references
5. Step 4: describe only new motion — do not repeat prior action
6. Step 5: seed, aspect ratio, resolution, fps, and color tone must match Segment 1 exactly
7. Submit Segment 2

Multi-Segment Iteration & Assembly

Repeat for Part 3 and beyond, always using the previous segment's last frame as the next scene reference. Assemble all clips in CapCut, Premiere, DaVinci Resolve, or FFmpeg; add 0.2–0.5s cross-dissolve at joins for smoother transitions.

Method C · Template + Fixed Seed Fast Sequel

If Segment 1 used a template, re-apply the same template via Use Template, then change only the project name, continuation prompt, motion description, and seed (matching Segment 1), plus upload the last-frame reference.

Method D · 30-Second Strategy for 15s VIP Users

Produce two 15-second linked segments (Part 1 + Part 2) and assemble in post for ~30 seconds total; or upgrade to SVIP / SVIP+ for native 30-second single-pass generation.

6. Generation History

History lists all submitted video jobs. Each entry includes:

• Project name — from Step 1, or a truncated video description
• Motion summary — excerpt from Step 4
• Status — queued / processing / completed / failed, etc.
• Submitted at — job creation timestamp

Video also provides a quick link to History. When building sequels, cross-reference Segment 1 parameters here.

7. VIP Membership & Plans

View the full plan matrix on VIP . Standard five-tier comparison:

Subscription Flow

1. Select a plan on VIP and click Subscribe
2. You will be taken to the Plan checkout for that plan automatically — no manual URL entry required
3. Complete USDT transfer and order submission per Section 8
4. After approval, verify expiry and quota on Profile

8. USDT Payment Workflow

Common Issues

• Invalid TxHash — must be 64 hexadecimal characters
• Duplicate TxHash — one on-chain transaction cannot bind multiple orders
• No deposit address — contact the site administrator
• Session expired — return to VIP and restart checkout

9. Account & Orders

9.1 Account Overview

Click your sidebar avatar to open Profile :

• Username, email, current membership role
• VIP expiry date (if subscribed)
• Remaining Video Quota for today

9.2 Order Management

My Orders tracks USDT subscription orders: order ID, plan, amount, TxHash, status, and timestamp. If membership has not activated, verify order status here first.

10. Parameter Reference

Prompt Structure

Use a four-part framework: subject + environment + mood + style. Aim for 50–300 words — specific enough to guide the model, concise enough to stay focused.

Weight Strategy

• Keep character, change background → lower scene weight, maintain high character weight
• Strict scene reproduction → scene weight 0.8+
• Sequel bridging → last frame as scene ref, weight 0.7–0.85

Camera Language

• Static — dialogue, portraits, stable composition
• Push / pull — emotional build or environment reveal
• Orbit — product or character showcase
• Track — walking, athletic motion
• Handheld — documentary feel, tension

Random Seed

• -1 — fully random; results not reproducible
• Fixed positive integer — reproducible when prompt and params match; strongly recommended for sequels

11. FAQ

Q: "Confirm & Submit" does nothing?

A: Non-VIP users trigger a subscription dialog. Go to VIP to subscribe.

Q: Can free users generate video?

A: Free users may explore the full wizard, but formal submission requires VIP or above.

Q: How do I continue after a 15-second clip?

A: See Section 5: extract last frame → create a new job with that reference → lock seed and output params → assemble in post.

Q: Is there a one-click sequel in history?

A: Not in the current release. Follow Section 5 to create linked jobs manually.

Q: Reference upload fails?

A: Check format (JPG/PNG/WebP), file size (< 10 MB recommended), and limits (3 scene / 4 character images).

Q: Job stuck in queue?

A: Peak-hour queuing is normal; VIP receives priority. Contact an administrator if status does not change for an extended period.

Q: Transferred on the wrong network?

A: Non-TRC20 transfers typically cannot be credited and are difficult to recover. Always verify the network in your wallet.

Q: What happens when VIP expires?

A: New submissions are blocked; history remains accessible. Renew to restore full access.

12. Best Practices

• Template-first workflow — start from a template; adjust description and motion only to reduce iteration cost
• Step-by-step validation — review the live preview panel after each step
• Archive sequel parameters — snapshot Step 6 before submit; reuse seed and output specs exactly
• Quota planning — segmented production consumes multiple daily slots; budget Part 1 / Part 2 tests accordingly
• Vertical content — 9:16 + 1080p for short-form platforms
• Cinematic look — 24 fps + 16:9 + warm tone + push/pull camera
• Iteration on failure — tweak motion description or seed first; avoid full parameter resets