Wavebreak Media
VIDEO DATASET

4K & HD AI-Generated Video Dataset for AI Training

Access 2,447 horizontal AI-generated video clips spanning 4K/UHD and HD, with approximately 4.4 hours of synthetic content. The mixed-resolution corpus is intended for high-resolution generation research, quality assessment, synthetic-content recognition, and pipelines that must operate consistently across resolution tiers.

A delivery manifest can distinguish 4K/UHD from HD assets and document resolution origin where known. Asset-level records can also state the generation workflow, provenance, synthetic-content flag, generation date, model or production source where disclosure is permitted, prompt availability, reference-input status, human-review state, and known limitations. High-resolution synthetic material of this kind is typically assessed alongside the wider generative AI training data Wavebreak Media licenses.

Request Full Dataset Browse Dataset Library
2,447
Clips
4.4
Total Hours
6.5s
Average Clip Length
4K + HD
Formats
Horizontal
Orientation
AI-Generated
Content Type

Dataset Preview

Representative horizontal previews from the mixed 4K/UHD and HD AI-generated corpus, showing variation in subjects, scenes, motion, detail, and synthetic artifacts.

The examples illustrate how generated content varies across scene complexity and resolution, including fine-detail retention, motion consistency, and visible generation artifacts.

Key Highlights

  • • Resolution labels suitable for mixed-resolution training and evaluation splits
  • • Resolution-origin status can identify native, derived, transcoded, or upscaled files where known
  • • Explicit synthetic-content flag at asset level
  • • Generation workflow, date, provenance, and production-source status can be supplied from retained records
  • • Prompt and reference-input availability are expressed as statuses rather than assumed to be complete
  • • Human-review state and known artifact or limitation notes can be included per clip

Metadata Fields

clip_id Unique identifier linking the media file to its manifest record
title Descriptive title of the AI-generated video clip
description Natural-language description of the visible generated content
keywords Keyword tags describing subjects, environments, objects, visual concepts, and scenes in the clip
resolution Video resolution, including 4K/UHD or HD
resolution_origin Native, derived, transcoded, or upscaled status where this can be established from retained production records
duration Length of the individual video clip
fps Frames per second
orientation Video orientation; horizontal in this dataset
generation_workflow Recorded workflow, such as text-to-video, image-to-video, video-to-video, or a multi-stage production process, where retained
provenance Available source, ownership, production, ingestion, and transformation history for the asset
synthetic_content Explicit boolean or controlled-value flag identifying the clip as AI-generated or otherwise synthetic
generation_date Generation or production date where retained in the source or production record
model_or_production_source Generating model, platform, studio, or production source where known and contractually permitted for disclosure
prompt_availability Status indicating whether a full prompt, partial prompt, prompt summary, or no prompt can be supplied
reference_input_status Status for image, video, or other reference inputs, including unavailable or restricted source material
human_review Whether technical, visual, policy, or metadata review was completed by a human
known_limitations Asset-level notes for identified issues such as temporal instability, anatomy errors, text artifacts, physics errors, or resolution processing

Technical Specifications

Dataset Type

Mixed-Resolution Synthetic Video: 4K/UHD + HD Horizontal

Primary Technical Intent

High-resolution evaluation, synthetic-content recognition, and resolution-robust processing

Clip Count

2,447 clips

Total Duration

4.4 hours

Total Duration in Minutes

Approximately 264.9 minutes

Average Clip Length

Approximately 6.5 seconds

Resolution

UHD: 3840 × 2160
DCI 4K: 4096 × 2160 where available
Full HD: 1920 × 1080

Resolution Distribution

4K/UHD and HD counts can be supplied in the delivery manifest for the selected corpus

Resolution Provenance

Native, derived, transcoded, or upscaled status recorded where known

Codec

H.264

4K/UHD Bitrate

Approximately 87–93 Mbps

HD Bitrate

Approximately 22 Mbps

Container

MP4

Orientation

Horizontal

Aspect Ratio Mixed horizontal formats

16:9 UHD/HD; 256:135 DCI 4K where available

Synthetic-Content Label

Explicit asset-level flag

Generation Documentation

Workflow, date, provenance, and model or production source supplied where retained and permitted

Prompt & Reference Inputs

Availability and restriction status recorded separately

Review & Limitations

Human-review status and known technical or visual limitations supplied where recorded

Licensing, Documentation, and Delivery

Wavebreak Media can provide applicable licensing, generation provenance, and disclosure status information for the mixed-resolution AI-generated video content, with dataset delivery organized around agreed file, metadata, and technical requirements. See our Dataset Licensing & Compliance and Dataset Delivery & Security pages for details on rights documentation, packaging, and secure transfer.

Need the Full Dataset?

Request the mixed 4K/UHD and HD synthetic video corpus with the required resolution split, provenance fields, generation documentation, review status, and limitation reporting.

Request Full Dataset Browse Dataset Library

Related Datasets

Video Datasets for AI Training

HD AI-Generated Video Dataset for AI Training

A standardized 1920 × 1080 horizontal synthetic-video corpus for scalable classification, retrieval, AI-video detection, and efficient model training.

View Dataset Video Datasets for AI Training

HD Vertical AI-Generated Video Dataset for AI Training

A 1080 × 1920 AI-generated video corpus for mobile generation, portrait composition, short-form outputs, and vertical-model evaluation.

View Dataset Video Datasets for AI Training

4K & HD Animation & Motion Graphics Dataset for AI Training

Mixed-resolution horizontal 2D, 3D, and motion graphics for high-resolution generation, compositing, motion understanding, and resolution-diverse training.

View Dataset

Selected Partners

Selected Wavebreak Media partners