目前最快LTX-Video文/图生视频模型，8G可玩

更新: 12/21/2024 字数: 0 字 时长: 0 分钟

概述

LTX-Video是第一个基于DiT的视频生成模型，可以实时生成高质量的视频，它可以以768x512的分辨率生成24 FPS视频，比观看它们所需的速度更快。该模型在多样化视频的大规模数据集上进行训练，可以生成内容逼真多样的高分辨率视频。

项目简介

项目信息

• 项目主页：地址
• 项目介绍：地址

在线体验

huggingface体验：点击访问

fal.ai文生视频体验：点击访问

fal.ai图生视频体验：点击访问

模型下载

• 通过HuggingFace：访问
• 通过HF镜像：访问

官方提示词生成规范(重要)

• 提示词尽可能规范，如果太简单或随意，显示的结果很差
• 提示词只支持英文，下面是官方推荐的提示词规范和要求，请仔细阅读

提示词规范

• 将下面提示词，发给聊天GPT类大语言对话模型，国内也可用豆包、通义千问、kimi、deepseek等

When writing prompts, focus on detailed, chronological descriptions of actions and scenes. Include specific movements, appearances, camera angles, and environmental details - all in a single flowing paragraph. Start directly with the action, and keep descriptions literal and precise. Think like a cinematographer describing a shot list. Keep within 200 words. For best results, build your prompts using this structure:

- Start with main action in a single sentence
- Add specific details about movements and gestures
- Describe character/object appearances precisely
- Include background and environment details
- Specify camera angles and movements
- Describe lighting and colors
- Note any changes or sudden events

你是一个提示词创作专家，我给出关键词，请根据上面提示词要求完善提示词；

提示词请以英文返回，如果明白，请回复明白；

下面是几个示例，请参考示例：

示例一：A woman with long brown hair and light skin smiles at another woman...
A woman with long brown hair and light skin smiles at another woman with long blonde hair. The woman with brown hair wears a black jacket and has a small, barely noticeable mole on her right cheek. The camera angle is a close-up, focused on the woman with brown hair's face. The lighting is warm and natural, likely from the setting sun, casting a soft glow on the scene. The scene appears to be real-life footage.

示例二：A woman walks away from a white Jeep parked on a city street at night...
A woman walks away from a white Jeep parked on a city street at night, then ascends a staircase and knocks on a door. The woman, wearing a dark jacket and jeans, walks away from the Jeep parked on the left side of the street, her back to the camera; she walks at a steady pace, her arms swinging slightly by her sides; the street is dimly lit, with streetlights casting pools of light on the wet pavement; a man in a dark jacket and jeans walks past the Jeep in the opposite direction; the camera follows the woman from behind as she walks up a set of stairs towards a building with a green door; she reaches the top of the stairs and turns left, continuing to walk towards the building; she reaches the door and knocks on it with her right hand; the camera remains stationary, focused on the doorway; the scene is captured in real-life footage

示例三：A clear, turquoise river flows through a rocky canyon...
A clear, turquoise river flows through a rocky canyon, cascading over a small waterfall and forming a pool of water at the bottom.The river is the main focus of the scene, with its clear water reflecting the surrounding trees and rocks. The canyon walls are steep and rocky, with some vegetation growing on them. The trees are mostly pine trees, with their green needles contrasting with the brown and gray rocks. The overall tone of the scene is one of peace and tranquility.

示例四：Two police officers in dark blue uniforms and matching hats...
Two police officers in dark blue uniforms and matching hats enter a dimly lit room through a doorway on the left side of the frame. The first officer, with short brown hair and a mustache, steps inside first, followed by his partner, who has a shaved head and a goatee. Both officers have serious expressions and maintain a steady pace as they move deeper into the room. The camera remains stationary, capturing them from a slightly low angle as they enter. The room has exposed brick walls and a corrugated metal ceiling, with a barred window visible in the background. The lighting is low-key, casting shadows on the officers' faces and emphasizing the grim atmosphere. The scene appears to be from a film or television show.

• 翻译

编写提示词时，请关注动作和场景的详细、按时间顺序的描述。包括特定的动作、外观、相机角度和环境细节——都在一个流畅的段落中。直接从动作开始，并保持描述的文字和精确。像电影摄影师描述镜头列表一样思考。保持在200字以内。为了获得最佳效果，使用以下结构构建提示：

从一个句子中的主要动作开始

添加有关动作和手势的特定详细信息

准确描述人物/物体的外观

包括背景和环境细节

指定摄像机角度和运动

描述灯光和颜色

注意任何变化或突发事件

查看示例以获取更多灵感。

• 有些大语言对话，服从性相对差点，我个人使用deepseek，按个人喜好和AI服从性选择

让AI按规范生成提示词

• 根据你的关键字，大语言AI返回完善后的提示词

实测

优点

• 生成速度相对目前市面上开源生成视频模型更快，显存占用更小，如果不满意可多抽几次卡
• 4090生成5s视频，大概30s（目前个人体验最快的生成视频模型）

• 下面的提示词，是根据提示词生成规范生成的

• 关键词：一个男人在办公室打电话

A man sits at his desk in a modern office, speaking on the phone with a focused expression. He wears a crisp white shirt and a navy blue tie, his sleeves rolled up to reveal his forearms. His desk is neatly organized, with a laptop, a notepad, and a cup of coffee arranged in a tidy manner. The camera angle is a medium shot, capturing him from the waist up, with the office environment visible in the background. The office is bright and spacious, with large windows letting in natural light, and a few colleagues working at their desks nearby. The lighting is soft and even, with the sunlight casting a warm glow on the scene. The man gestures slightly with his free hand as he speaks, his movements precise and deliberate. The camera remains stationary, focusing on his calm demeanor and the subtle details of his office surroundings. The background noise is minimal, with the faint hum of computers and the occasional murmur of conversation adding to the realistic atmosphere. The scene is captured in real-time, emphasizing the man's professionalism and the efficiency of the workplace.

翻译：现代办公室里，一名男子坐在办公桌前，神情专注地讲着电话。他穿着挺括的白衬衫，打着藏青色的领带，袖子卷起露出前臂。他的办公桌整齐地摆放着一台笔记本电脑、一个记事本和一杯咖啡，摆放得整整齐齐。相机角度为中景拍摄，从腰部以上捕捉他，背景中可以看到办公环境。办公室明亮宽敞，大窗户让自然光照进来，几名同事在附近的办公桌前工作。灯光柔和均匀，阳光在现场投下温暖的光芒。 男子说话时用空着的手略作手势，动作精准而深思熟虑。摄像机保持静止，聚焦在他平静的举止和办公室周围微妙的细节上。背景噪声极小，电脑的微弱嗡嗡声和偶尔的谈话杂音增加了现实的气氛。现场被实时捕捉，强调了男子的专业精神和工作场所的效率。

• 关键词：巨型机器人变形金刚汽车人从海洋中浮现，海浪撞击着它的金属框架。场景显示，当机器人发光的眼睛扫视地平线时，惊恐的海滩游客正在逃离

A colossal Transformer, the Autobot, emerges from the ocean, its massive metallic frame towering over the crashing waves. The robot, with intricate mechanical details and glowing blue eyes, rises from the depths, its transformation sequence unfolding in slow motion as it shifts from vehicle to robot mode. The camera angle is a wide shot, capturing the entire scene from a low vantage point, with the ocean stretching out behind the towering figure. The waves crash violently against its metallic body, spraying water in all directions. On the beach, panicked tourists in swimsuits and beachwear flee in various directions, their movements chaotic and frantic. The camera shifts to a close-up of the Autobot's glowing eyes as they scan the horizon, their intensity adding to the sense of impending danger. The lighting is dramatic, with the sun setting in the background, casting a fiery orange glow on the water and the robot's metallic surface. The scene is tense and action-packed, with the sound of roaring waves and distant screams adding to the chaos. The camera slowly zooms in on the Autobot's face, focusing on its determined expression as it prepares for action.

一个巨大的Transformer，汽车人，从海洋中浮现出来，它巨大的金属框架高耸在汹涌的海浪之上。这个机器人，有着复杂的机械细节和发光的蓝眼睛，从深处升起，它的变换序列在从车辆模式转变为机器人模式时以慢动作展开。相机角度是一个广角镜头，从低的有利位置捕捉整个场景，海洋在高耸的人物后面伸展开来。海浪猛烈撞击着它的金属身体，向四面八方喷水。在海滩上，惊慌失措的游客穿着泳衣和沙滩装向各个方向逃离，他们的动作混乱而疯狂。 当汽车人扫视地平线时，镜头转向了他们发光眼睛的特写镜头，它们的强度增加了危险即将到来的感觉。灯光非常引人注目，背景是夕阳，在水面和机器人的金属表面投下炽热的橙色光芒。场景紧张而充满动作，咆哮的海浪声和远处的尖叫声加剧了混乱。镜头慢慢放大汽车人的脸，专注于它准备行动时坚定的表情。

免安装环境win-webui版💎

预览

说明

• 软件已经过测试，测试平台为Windows10和Nvidia-4090显卡
• 不支持AMD显卡及核显，显存尽量8GB以上，cuda-12版本，低显存或低cuda版本不保证正常使用
• 点此查看自己的显卡相关信息
• 压缩包已包含依赖的环境模型等大文件，无需安装环境，点开即用；
• 大小：20GB

下载地址

百度网盘

夸克网盘

秋叶ComfyUI便携版🔑

• 秋叶ComfyUI基础教程：查看

注意

注意

下面两种安装方式和模型只为记录安装过程，直接从网盘下载后启动即可，无需再次安装

使用安装管理器方式

• 输入安装地址：https://github.com/Lightricks/ComfyUI-LTXVideo

• 安装成功后，重启ComfyUI

使用git仓库手动方式

• 如果安装管理安装有问题，可以使用手动方式
• 克隆仓库

# 进入comfyUI节点目录
cd custom_nodes

git clone https://github.com/Lightricks/ComfyUI-LTXVideo

cd ComfyUI-LTXVideo

• 使用当前ComyUI环境的python进行安装(注意：不要直接使用python.exe)

# 当前ComfyUI环境的python
..\..\python\python.exe -m pip install -r requirements.txt

安装成功后，重启ComfyUI

模型下载

1. 下载ltx-video-2b-v0.9.safetensors 模型到 models/checkpoints.
2. 安装 git-lfs ，克隆或者下载模型到： models/text_encoders:

# 如果使用git下载太慢，可以先git clone，会自动下载小文件，
# 一直下载不动或太慢时，可以先`ctrl+c`断开，然后手动将大文件下载下来
cd models/text_encoders 

git lfs linstall

# 国内可使用：git clone https://hf-mirror.con/PixArt-alpha/PixArt-XL-2-1024-MS
git clone https://huggingface.co/PixArt-alpha/PixArt-XL-2-1024-MS

下载地址

建议

说明：

• 已安装相关依赖以及工作流

• 为避免单个压缩包体积过大，几个大的模型分开单独文件夹上传到网盘

百度网盘

夸克网盘

使用

• 打开http://127.0.0.1:8188/
• 将工作流拖到打开的web页面里

STG优化生成质量

• webui版只为快速试用，暂未添加STG优化

• 结果