WAN 3.0 Pro Uncensored Video Model

Overview

WAN 3.0 Pro Uncensored generates a single video clip from a written shot description. Wiro-Partners provides this as a Pro configuration of the Wan 3.0 family for higher-fidelity outputs and fewer refusals.

You can generate from text alone, or anchor the motion to a still first frame. You can also add a still last frame to guide where the shot ends. The model returns a ready-to-use MP4 clip, which helps you prototype motion scenes without filming.

What you can build

• Text-to-video scene drafts for storyboards and previsualization
• Social clips for Reels, Shorts, and TikTok from a single shot brief
• Product hero shots with controlled opening and closing frames
• Character or outfit continuity tests using a fixed first frame
• Music-video style visuals where you describe both visuals and sound
• Iteration workflows where you keep the same setup and vary the seed

Inputs

• A required prompt in plain text that describes one continuous shot, including subject, action, camera movement, lighting, and sound if you want audio.
• An optional first frame image that becomes the opening frame of the video.
• An optional last frame image that guides the final frame of the video.
• A required duration selection of 5, 10, 15, 20, 25, or 30 seconds.
• A required quality tier selection: 1080p, 2K, or 4K.
• An optional aspect ratio selection. You can set 16:9, 4:3, 1:1, 3:4, or 9:16, or leave it on adaptive so it derives from your prompt and any frame images.
• An optional toggle to include an audio track in the output.
• An optional toggle that rewrites and expands your prompt for more detail. This is meant to help short prompts.
• An optional numeric seed for reproducibility. Use an integer from 0 to 2,147,483,647.

Outputs

• One MP4 video clip at your selected duration and quality tier.
• Framing follows your chosen aspect ratio, or adapts when you choose adaptive.
• If you enable audio, the MP4 includes an audio track generated for the scene you described.

Recommended settings

• If your prompt is short, enable prompt expansion so the model adds missing shot details.
• If you need repeatable iterations, set a fixed seed and change only one prompt detail per run.
• If you need tighter continuity, provide a first frame. Add a last frame when the ending pose or composition matters.
• If audio matters, describe it explicitly in the prompt. Mention dialogue, ambience, and timing cues.

Limitations

• The clip length is limited to the supported duration options, up to 30 seconds.
• This listing focuses on generation from text plus optional first and last frames. It does not act as a video editor for an existing input clip.
• The interface does not provide a dedicated “do not include” field, so exclusions must be written into the main prompt.
• Low-quality or ambiguous inputs can degrade results. Blurry frames, heavy compression, screenshots with overlays, or collages can cause unstable motion and artifacts.
• “Uncensored” configurations can still fail on some requests depending on platform and legal constraints.

Safety & compliance

• Don’t include private information in prompts or uploads.
• Only upload images you own or have rights to use.
• Don’t generate non-consensual sexual content, sexual content involving minors, or violent extremist content.
• Don’t create deceptive impersonations or deepfakes of real people.
• If you’re building a user-facing app, add your own moderation and abuse controls. Uncensored endpoints shift more responsibility to you.