ltx-2-19b-dev-fp8AI 音视频生成模型：

LTX-2-19B-dev-FP8 模型典型应用场景

1. 短视频与自媒体创作

• 文字直接生成剧情短片、风景动态视频、搞笑段子、知识科普动画
• 单张图片生成动态壁纸、人物微动、场景动效
• 快速批量产出抖音、快手、小红书竖屏短视频

2. 广告与营销物料

• 品牌宣传短片、产品展示动态视频
• 电商主图视频、商品动态演示
• 活动海报→一键生成动态海报视频

3. 影视 / 动画前期预演

• 剧本分镜快速可视化
• 概念短片、预告片小样生成
• 场景、特效、镜头运镜的低成本测试

4. 游戏与虚拟内容

• 游戏宣传动态 PV、场景漫游视频
• NPC 动态展示、剧情过场动画初稿
• 虚拟主播背景动态视频

5. 教育与培训内容

• 课件动态演示、微课视频
• 安全培训、操作流程演示视频
• 无需拍摄即可制作教学短片

6. 个人创意与 AIGC 创作

• 小说 / 故事可视化动态短片
• 照片复活、老照片动态化
• 音乐可视化、随节奏生成画面

7. 设计与视觉行业

• 平面设计稿→动态视觉效果
• 室内 / 建筑效果图生成漫游视频
• 插画、原画转为动态短片

它最突出的优势场景

1. 需要较长视频（相对普通文生视频模型）
2. 追求画质接近真实 / 电影感
3. 希望音画一起生成，不用后期配音效
4. 本地部署、不想依赖云端 API
5. 显存有限但想跑大参数量视频模型（FP8 量化优势）

安装说明：

一、先准备好硬件

• 显卡显存：≥16GB（24GB 更稳）
• 内存：≥32GB（推荐 64GB）
• 硬盘：至少 50GB 空闲 SSD

二、安装插件（必须）

1. 安装 ComfyUI-LTXVideo 插件
2. 重启 ComfyUI

custom_nodes/ComfyUI-LTXVideo

三、模型文件放哪里

1. LTX-2 19B FP8 主模型

models/ltx_video/ltx-2-19b-dev-fp8.safetensors

2. VAE

models/vae/ltx_vae.safetensors

3. T5 文本编码器

models/clip/t5xxl_fp8_e4m3.safetensors

4. 音频模型（可选）

models/audio_models/lm_small_v1.safetensors
models/audio_models/vocoder_v1.safetensors

四、直接可用的 ComfyUI 工作流（复制即用）

{"last_node_id":29,"nodes":[{"id":1,"type":"LTXLoadModel","pos":[29,85],"size":[322,170],"flags":{},"order":0,"mode":0,"inputs":[],"outputs":[{"name":"MODEL","type":"MODEL","links":[2],"shape":3},{"name":"VAE","type":"VAE","links":[4],"shape":3}],"properties":{"Node name for S&R":"LTXLoadModel"},"widgets_values":["ltx-2-19b-dev-fp8.safetensors","ltx_vae.safetensors"]},{"id":2,"type":"LTXLoadTextEncoder","pos":[399,84],"size":[322,130],"flags":{},"order":1,"mode":0,"inputs":[],"outputs":[{"name":"TEXT_ENCODER","type":"CLIP","links":[3],"shape":3}],"properties":{"Node name for S&R":"LTXLoadTextEncoder"},"widgets_values":["t5xxl_fp8_e4m3.safetensors"]},{"id":3,"type":"LTXTextEncode","pos":[761,85],"size":[322,130],"flags":{},"order":2,"mode":0,"inputs":[{"name":"text_encoder","type":"CLIP","link":3}],"outputs":[{"name":"CONDITIONING","type":"CONDITIONING","links":[5],"shape":3}],"properties":{"Node name for S&R":"LTXTextEncode"},"widgets_values":["A beautiful landscape, cinematic, 4K, detailed, realistic, smooth motion"]},{"id":4,"type":"LTXEmptyLatentVideo","pos":[28,291],"size":[240,100],"flags":{},"order":3,"mode":0,"inputs":[],"outputs":[{"name":"LATENT","type":"LATENT","links":[6],"shape":3}],"properties":{"Node name for S&R":"LTXEmptyLatentVideo"},"widgets_values":[768,512,16]},{"id":5,"type":"LTXSampler","pos":[397,289],"size":[324,260],"flags":{},"order":4,"mode":0,"inputs":[{"name":"model","type":"MODEL","link":2},{"name":"vae","type":"VAE","link":4},{"name":"positive","type":"CONDITIONING","link":5},{"name":"latent_video","type":"LATENT","link":6}],"outputs":[{"name":"LATENT","type":"LATENT","links":[7],"shape":3}],"properties":{"Node name for S&R":"LTXSampler"},"widgets_values":["euler","normal",14,6.5,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,"fixed","disable"]},{"id":6,"type":"VTAEDecode","pos":[398,579],"size":[240,100],"flags":{},"order":5,"mode":0,"inputs":[{"name":"vae","type":"VAE","link":4},{"name":"samples","type":"LATENT","link":7}],"outputs":[{"name":"VIDEO","type":"IMAGE","links":[8],"shape":3}],"properties":{"Node name for S&R":"VTAEDecode"},"widgets_values":["default"]},{"id":7,"type":"PreviewVideo","pos":[700,580],"size":[240,260],"flags":{},"order":6,"mode":0,"inputs":[{"name":"video","type":"IMAGE","link":8}],"outputs":[],"properties":{"Node name for S&R":"PreviewVideo"},"widgets_values":["ltx_output","video/mp4",1]}],"links":[[2,1,0,5,0],[3,2,0,3,0],[4,1,1,5,1],[5,3,0,5,2],[6,4,0,5,3],[7,5,0,6,1],[8,6,0,7,0]],"groups":[],"config":{},"extra":{},"version":0.4}

五、关键参数说明

• 分辨率：
  • 16G 显存：768×512
  • 24G 显存：896×512 / 1024×576
• 帧数：16 帧最稳（≈2 秒）
• CFG：6.5
• 步数：14–20

六、显存不够的解决办法

1. 开启 FP8
2. 分辨率降到 640×448
3. 帧数改为 8 帧
4. 关闭 highres fix
5. 关闭其他占用显存的程序

一、中文提示词模板（电影感 / 古风 / 写实 / 游戏）

1. 电影感大片

宏大电影场景，史诗级光影，大气运镜，细腻皮肤质感，真实布料纹理，4K超高清，动态流畅，专业摄影，浅景深，cinematic，sharp focus，masterpiece，best quality

2. 古风仙侠

古风仙侠女子，飘逸白衣，长发随风飘动，云雾缭绕山间，花瓣飞舞，水墨国风，柔和光影，仙气飘飘，唯美意境，流畅动态，东方美学，精致细节，4K

3. 写实生活 / 人物

真实人物，自然表情，细腻皮肤，自然光，窗边柔和光线，真实质感，流畅微表情，自然动作，生活化场景，高清细节，realistic，photorealistic

4. 游戏 CG / 科幻

科幻未来城市，飞行器穿梭，霓虹灯光，赛博朋克，金属质感，粒子特效，宏大场景，流畅运镜，电影级CG，high detail，unreal engine 5

负面通用（必加）

模糊，失真，抖动，变形，低清，马赛克，文字，水印，丑陋，畸形，扭曲，多余肢体，坏脸，坏手，lowres，blurry，ugly，disfigured

二、图生视频工作流（ComfyUI 直接粘贴）

{"last_node_id":35,"nodes":[{"id":1,"type":"LTXLoadModel","pos":[49,69],"size":[322,170],"flags":{},"order":0,"mode":0,"inputs":[],"outputs":[{"name":"MODEL","type":"MODEL","links":[2],"shape":3},{"name":"VAE","type":"VAE","links":[4],"shape":3}],"properties":{"Node name for S&R":"LTXLoadModel"},"widgets_values":["ltx-2-19b-dev-fp8.safetensors","ltx_vae.safetensors"]},{"id":2,"type":"LTXLoadTextEncoder","pos":[419,68],"size":[322,130],"flags":{},"order":1,"mode":0,"inputs":[],"outputs":[{"name":"TEXT_ENCODER","type":"CLIP","links":[3],"shape":3}],"properties":{"Node name for S&R":"LTXLoadTextEncoder"},"widgets_values":["t5xxl_fp8_e4m3.safetensors"]},{"id":3,"type":"LTXTextEncode","pos":[781,69],"size":[322,140],"flags":{},"order":2,"mode":0,"inputs":[{"name":"text_encoder","type":"CLIP","link":3}],"outputs":[{"name":"CONDITIONING","type":"CONDITIONING","links":[5],"shape":3}],"properties":{"Node name for S&R":"LTXTextEncode"},"widgets_values":["A beautiful landscape, cinematic, 4K, detailed, realistic, smooth motion"]},{"id":4,"type":"LoadImage","pos":[39,261],"size":[310,280],"flags":{},"order":3,"mode":0,"inputs":[],"outputs":[{"name":"IMAGE","type":"IMAGE","links":[6],"shape":3}],"properties":{"Node name for S&R":"LoadImage"},"widgets_values":["【替换你的图片】","image"]},{"id":5,"type":"LTXImageToVideoLatent","pos":[397,260],"size":[324,120],"flags":{},"order":4,"mode":0,"inputs":[{"name":"IMAGE","type":"IMAGE","link":6}],"outputs":[{"name":"LATENT","type":"LATENT","links":[7],"shape":3}],"properties":{"Node name for S&R":"LTXImageToVideoLatent"},"widgets_values":[16]},{"id":6,"type":"LTXSampler","pos":[398,409],"size":[324,260],"flags":{},"order":5,"mode":0,"inputs":[{"name":"model","type":"MODEL","link":2},{"name":"vae","type":"VAE","link":4},{"name":"positive","type":"CONDITIONING","link":5},{"name":"latent_video","type":"LATENT","link":7}],"outputs":[{"name":"LATENT","type":"LATENT","links":[8],"shape":3}],"properties":{"Node name for S&R":"LTXSampler"},"widgets_values":["euler","normal",14,6.5,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,"fixed","disable"]},{"id":7,"type":"VTAEDecode","pos":[399,699],"size":[240,100],"flags":{},"order":6,"mode":0,"inputs":[{"name":"vae","type":"VAE","link":4},{"name":"samples","type":"LATENT","link":8}],"outputs":[{"name":"VIDEO","type":"IMAGE","links":[9],"shape":3}],"properties":{"Node name for S&R":"VTAEDecode"},"widgets_values":["default"]},{"id":8,"type":"PreviewVideo","pos":[701,700],"size":[240,260],"flags":{},"order":7,"mode":0,"inputs":[{"name":"video","type":"IMAGE","link":9}],"outputs":[],"properties":{"Node name for S&R":"PreviewVideo"},"widgets_values":["ltx_img2vid","video/mp4",1]}],"links":[[2,1,0,6,0],[3,2,0,3,0],[4,1,1,6,1],[5,3,0,6,2],[6,4,0,5,0],[7,5,0,6,3],[8,6,0,7,1],[9,7,0,8,0]],"groups":[],"config":{},"extra":{},"version":0.4}

三、带音频生成的完整音视频工作流（最终完整版）

{"last_node_id":45,"nodes":[{"id":1,"type":"LTXLoadModel","pos":[39,49],"size":[322,170],"flags":{},"order":0,"mode":0,"inputs":[],"outputs":[{"name":"MODEL","type":"MODEL","links":[2],"shape":3},{"name":"VAE","type":"VAE","links":[4],"shape":3}],"properties":{"Node name for S&R":"LTXLoadModel"},"widgets_values":["ltx-2-19b-dev-fp8.safetensors","ltx_vae.safetensors"]},{"id":2,"type":"LTXLoadTextEncoder","pos":[409,48],"size":[322,130],"flags":{},"order":1,"mode":0,"inputs":[],"outputs":[{"name":"TEXT_ENCODER","type":"CLIP","links":[3],"shape":3}],"properties":{"Node name for S&R":"LTXLoadTextEncoder"},"widgets_values":["t5xxl_fp8_e4m3.safetensors"]},{"id":3,"type":"LTXTextEncode","pos":[771,49],"size":[322,140],"flags":{},"order":2,"mode":0,"inputs":[{"name":"text_encoder","type":"CLIP","link":3}],"outputs":[{"name":"CONDITIONING","type":"CONDITIONING","links":[5],"shape":3}],"properties":{"Node name for S&R":"LTXTextEncode"},"widgets_values":["Cinematic scene, realistic, smooth motion, detailed atmosphere"]},{"id":4,"type":"LTXEmptyLatentVideo","pos":[38,241],"size":[240,100],"flags":{},"order":3,"mode":0,"inputs":[],"outputs":[{"name":"LATENT","type":"LATENT","links":[6],"shape":3}],"properties":{"Node name for S&R":"LTXEmptyLatentVideo"},"widgets_values":[768,512,16]},{"id":5,"type":"LTXSampler","pos":[407,240],"size":[324,260],"flags":{},"order":4,"mode":0,"inputs":[{"name":"model","type":"MODEL","link":2},{"name":"vae","type":"VAE","link":4},{"name":"positive","type":"CONDITIONING","link":5},{"name":"latent_video","type":"LATENT","link":6}],"outputs":[{"name":"LATENT","type":"LATENT","links":[7],"shape":3}],"properties":{"Node name for S&R":"LTXSampler"},"widgets_values":["euler","normal",14,6.5,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,"fixed","disable"]},{"id":6,"type":"VTAEDecode","pos":[408,529],"size":[240,100],"flags":{},"order":5,"mode":0,"inputs":[{"name":"vae","type":"VAE","link":4},{"name":"samples","type":"LATENT","link":7}],"outputs":[{"name":"VIDEO","type":"IMAGE","links":[8],"shape":3}],"properties":{"Node name for S&R":"VTAEDecode"},"widgets_values":["default"]},{"id":9,"type":"LTXAudioLM","pos":[38,620],"size":[310,140],"flags":{},"order":6,"mode":0,"inputs":[],"outputs":[{"name":"AUDIO_LM","type":"AUDIO_MODEL","links":[10],"shape":3}],"properties":{"Node name for S&R":"LTXAudioLM"},"widgets_values":["lm_small_v1.safetensors"]},{"id":10,"type":"LTXVocoder","pos":[39,770],"size":[310,140],"flags":{},"order":7,"mode":0,"inputs":[],"outputs":[{"name":"VOCODER","type":"AUDIO_MODEL","links":[11],"shape":3}],"properties":{"Node name for S&R":"LTXVocoder"},"widgets_values":["vocoder_v1.safetensors"]},{"id":11,"type":"LTXGenerateAudio","pos":[408,640],"size":[310,140],"flags":{},"order":8,"mode":0,"inputs":[{"name":"AUDIO_LM","type":"AUDIO_MODEL","link":10},{"name":"VOCODER","type":"AUDIO_MODEL","link":11}],"outputs":[{"name":"AUDIO","type":"AUDIO","links":[12],"shape":3}],"properties":{"Node name for S&R":"LTXGenerateAudio"},"widgets_values":["atmospheric cinematic background music, calm, immersive",16]},{"id":12,"type":"PreviewVideoAudio","pos":[771,530],"size":[280,300],"flags":{},"order":9,"mode":0,"inputs":[{"name":"video","type":"IMAGE","link":8},{"name":"audio","type":"AUDIO","link":12}],"outputs":[],"properties":{"Node name for S&R":"PreviewVideoAudio"},"widgets_values":["ltx_av_output","video/mp4",1]}],"links":[[2,1,0,5,0],[3,2,0,3,0],[4,1,1,5,1],[5,3,0,5,2],[6,4,0,5,3],[7,5,0,6,1],[8,6,0,12,0],[10,9,0,11,0],[11,10,0,11,1],[12,11,0,12,1]],"groups":[],"config":{},"extra":{},"version":0.4}

四、使用要点（16GB 显存必看）

• 分辨率：768×512
• 帧数：16 帧（约 2 秒）
• 步数：14 步
• CFG：6.5
• 模型必须放对路径：
  • models/ltx_video/ltx-2-19b-dev-fp8.safetensors
  • models/clip/t5xxl_fp8_e4m3.safetensors
  • models/vae/ltx_vae.safetensors
  • 音频：models/audio_models/

• 超低显存优化版（12GB 也能跑）

• 一、12GB 显存也能跑的核心参数（必设）

  • 分辨率：640 × 384
  • 帧数：8 帧（≈1 秒）
  • 采样步数：12 步
  • CFG：5.0
  • 采样器：euler
  • 调度：normal

  16GB 显存可以用：704 × 400 / 10 帧

  ---

  二、ComfyUI 启动参数（关键！省显存）

  启动 ComfyUI 时加上这些参数：
  plaintext

  --fp8-upcast
  --lowvram
  --cpu-offload
  --dont-upcast-attention
  
  能再省 2～3GB 显存。

  ---

  三、超低显存专用工作流（复制即用）

  plaintext

  {"last_node_id":30,"nodes":[{"id":1,"type":"LTXLoadModel","pos":[29,85],"size":[322,170],"flags":{},"order":0,"mode":0,"inputs":[],"outputs":[{"name":"MODEL","type":"MODEL","links":[2],"shape":3},{"name":"VAE","type":"VAE","links":[4],"shape":3}],"properties":{"Node name for S&R":"LTXLoadModel"},"widgets_values":["ltx-2-19b-dev-fp8.safetensors","ltx_vae.safetensors"]},{"id":2,"type":"LTXLoadTextEncoder","pos":[399,84],"size":[322,130],"flags":{},"order":1,"mode":0,"inputs":[],"outputs":[{"name":"TEXT_ENCODER","type":"CLIP","links":[3],"shape":3}],"properties":{"Node name for S&R":"LTXLoadTextEncoder"},"widgets_values":["t5xxl_fp8_e4m3.safetensors"]},{"id":3,"type":"LTXTextEncode","pos":[761,85],"size":[322,130],"flags":{},"order":2,"mode":0,"inputs":[{"name":"text_encoder","type":"CLIP","link":3}],"outputs":[{"name":"CONDITIONING","type":"CONDITIONING","links":[5],"shape":3}],"properties":{"Node name for S&R":"LTXTextEncode"},"widgets_values":["A beautiful landscape, cinematic, 4K, detailed, realistic, smooth motion"]},{"id":4,"type":"LTXEmptyLatentVideo","pos":[28,291],"size":[240,100],"flags":{},"order":3,"mode":0,"inputs":[],"outputs":[{"name":"LATENT","type":"LATENT","links":[6],"shape":3}],"properties":{"Node name for S&R":"LTXEmptyLatentVideo"},"widgets_values":[640,384,8]},{"id":5,"type":"LTXSampler","pos":[397,289],"size":[324,260],"flags":{},"order":4,"mode":0,"inputs":[{"name":"model","type":"MODEL","link":2},{"name":"vae","type":"VAE","link":4},{"name":"positive","type":"CONDITIONING","link":5},{"name":"latent_video","type":"LATENT","link":6}],"outputs":[{"name":"LATENT","type":"LATENT","links":[7],"shape":3}],"properties":{"Node name for S&R":"LTXSampler"},"widgets_values":["euler","normal",12,5.0,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,"fixed","enable"]},{"id":6,"type":"VTAEDecode","pos":[398,579],"size":[240,100],"flags":{},"order":5,"mode":0,"inputs":[{"name":"vae","type":"VAE","link":4},{"name":"samples","type":"LATENT","link":7}],"outputs":[{"name":"VIDEO","type":"IMAGE","links":[8],"shape":3}],"properties":{"Node name for S&R":"VTAEDecode"},"widgets_values":["default"]},{"id":7,"type":"PreviewVideo","pos":[700,580],"size":[240,260],"flags":{},"order":6,"mode":0,"inputs":[{"name":"video","type":"IMAGE","link":8}],"outputs":[],"properties":{"Node name for S&R":"PreviewVideo"},"widgets_values":["ltx_lowvram","video/mp4",1]}],"links":[[2,1,0,5,0],[3,2,0,3,0],[4,1,1,5,1],[5,3,0,5,2],[6,4,0,5,3],[7,5,0,6,1],[8,6,0,7,0]],"groups":[],"config":{},"extra":{},"version":0.4}
  

  ---

  四、图生视频低显存版

  plaintext

  {"last_node_id":36,"nodes":[{"id":1,"type":"LTXLoadModel","pos":[49,69],"size":[322,170],"flags":{},"order":0,"mode":0,"inputs":[],"outputs":[{"name":"MODEL","type":"MODEL","links":[2],"shape":3},{"name":"VAE","type":"VAE","links":[4],"shape":3}],"properties":{"Node name for S&R":"LTXLoadModel"},"widgets_values":["ltx-2-19b-dev-fp8.safetensors","ltx_vae.safetensors"]},{"id":2,"type":"LTXLoadTextEncoder","pos":[419,68],"size":[322,130],"flags":{},"order":1,"mode":0,"inputs":[],"outputs":[{"name":"TEXT_ENCODER","type":"CLIP","links":[3],"shape":3}],"properties":{"Node name for S&R":"LTXLoadTextEncoder"},"widgets_values":["t5xxl_fp8_e4m3.safetensors"]},{"id":3,"type":"LTXTextEncode","pos":[781,69],"size":[322,140],"flags":{},"order":2,"mode":0,"inputs":[{"name":"text_encoder","type":"CLIP","link":3}],"outputs":[{"name":"CONDITIONING","type":"CONDITIONING","links":[5],"shape":3}],"properties":{"Node name for S&R":"LTXTextEncode"},"widgets_values":["A beautiful landscape, cinematic, 4K, detailed, realistic, smooth motion"]},{"id":4,"type":"LoadImage","pos":[39,261],"size":[310,280],"flags":{},"order":3,"mode":0,"inputs":[],"outputs":[{"name":"IMAGE","type":"IMAGE","links":[6],"shape":3}],"properties":{"Node name for S&R":"LoadImage"},"widgets_values":["your_image.png","image"]},{"id":5,"type":"LTXImageToVideoLatent","pos":[397,260],"size":[324,120],"flags":{},"order":4,"mode":0,"inputs":[{"name":"IMAGE","type":"IMAGE","link":6}],"outputs":[{"name":"LATENT","type":"LATENT","links":[7],"shape":3}],"properties":{"Node name for S&R":"LTXImageToVideoLatent"},"widgets_values":[8]},{"id":6,"type":"LTXSampler","pos":[398,409],"size":[324,260],"flags":{},"order":5,"mode":0,"inputs":[{"name":"model","type":"MODEL","link":2},{"name":"vae","type":"VAE","link":4},{"name":"positive","type":"CONDITIONING","link":5},{"name":"latent_video","type":"LATENT","link":7}],"outputs":[{"name":"LATENT","type":"LATENT","links":[8],"shape":3}],"properties":{"Node name for S&R":"LTXSampler"},"widgets_values":["euler","normal",12,5.0,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,"fixed","enable"]},{"id":7,"type":"VTAEDecode","pos":[399,699],"size":[240,100],"flags":{},"order":6,"mode":0,"inputs":[{"name":"vae","type":"VAE","link":4},{"name":"samples","type":"LATENT","link":8}],"outputs":[{"name":"VIDEO","type":"IMAGE","links":[9],"shape":3}],"properties":{"Node name for S&R":"VTAEDecode"},"widgets_values":["default"]},{"id":8,"type":"PreviewVideo","pos":[701,700],"size":[240,260],"flags":{},"order":7,"mode":0,"inputs":[{"name":"video","type":"IMAGE","link":9}],"outputs":[],"properties":{"Node name for S&R":"PreviewVideo"},"widgets_values":["ltx_img2vid_lowvram","video/mp4",1]}],"links":[[2,1,0,6,0],[3,2,0,3,0],[4,1,1,6,1],[5,3,0,6,2],[6,4,0,5,0],[7,5,0,6,3],[8,6,0,7,1],[9,7,0,8,0]],"groups":[],"config":{},"extra":{},"version":0.4}
  

  ---

  五、带音频的低显存版

  plaintext

  {"last_node_id":46,"nodes":[{"id":1,"type":"LTXLoadModel","pos":[39,49],"size":[322,170],"flags":{},"order":0,"mode":0,"inputs":[],"outputs":[{"name":"MODEL","type":"MODEL","links":[2],"shape":3},{"name":"VAE","type":"VAE","links":[4],"shape":3}],"properties":{"Node name for S&R":"LTXLoadModel"},"widgets_values":["ltx-2-19b-dev-fp8.safetensors","ltx_vae.safetensors"]},{"id":2,"type":"LTXLoadTextEncoder","pos":[409,48],"size":[322,130],"flags":{},"order":1,"mode":0,"inputs":[],"outputs":[{"name":"TEXT_ENCODER","type":"CLIP","links":[3],"shape":3}],"properties":{"Node name for S&R":"LTXLoadTextEncoder"},"widgets_values":["t5xxl_fp8_e4m3.safetensors"]},{"id":3,"type":"LTXTextEncode","pos":[771,49],"size":[322,140],"flags":{},"order":2,"mode":0,"inputs":[{"name":"text_encoder","type":"CLIP","link":3}],"outputs":[{"name":"CONDITIONING","type":"CONDITIONING","links":[5],"shape":3}],"properties":{"Node name for S&R":"LTXTextEncode"},"widgets_values":["Cinematic scene, realistic, smooth motion, detailed atmosphere"]},{"id":4,"type":"LTXEmptyLatentVideo","pos":[38,241],"size":[240,100],"flags":{},"order":3,"mode":0,"inputs":[],"outputs":[{"name":"LATENT","type":"LATENT","links":[6],"shape":3}],"properties":{"Node name for S&R":"LTXEmptyLatentVideo"},"widgets_values":[640,384,8]},{"id":5,"type":"LTXSampler","pos":[407,240],"size":[324,260],"flags":{},"order":4,"mode":0,"inputs":[{"name":"model","type":"MODEL","link":2},{"name":"vae","type":"VAE","link":4},{"name":"positive","type":"CONDITIONING","link":5},{"name":"latent_video","type":"LATENT","link":6}],"outputs":[{"name":"LATENT","type":"LATENT","links":[7],"shape":3}],"properties":{"Node name for S&R":"LTXSampler"},"widgets_values":["euler","normal",12,5.0,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,"fixed","enable"]},{"id":6,"type":"VTAEDecode","pos":[408,529],"size":[240,100],"flags":{},"order":5,"mode":0,"inputs":[{"name":"vae","type":"VAE","link":4},{"name":"samples","type":"LATENT","link":7}],"outputs":[{"name":"VIDEO","type":"IMAGE","links":[8],"shape":3}],"properties":{"Node name for S&R":"VTAEDecode"},"widgets_values":["default"]},{"id":9,"type":"LTXAudioLM","pos":[38,620],"size":[310,140],"flags":{},"order":6,"mode":0,"inputs":[],"outputs":[{"name":"AUDIO_LM","type":"AUDIO_MODEL","links":[10],"shape":3}],"properties":{"Node name for S&R":"LTXAudioLM"},"widgets_values":["lm_small_v1.safetensors"]},{"id":10,"type":"LTXVocoder","pos":[39,770],"size":[310,140],"flags":{},"order":7,"mode":0,"inputs":[],"outputs":[{"name":"VOCODER","type":"AUDIO_MODEL","links":[11],"shape":3}],"properties":{"Node name for S&R":"LTXVocoder"},"widgets_values":["vocoder_v1.safetensors"]},{"id":11,"type":"LTXGenerateAudio","pos":[408,640],"size":[310,140],"flags":{},"order":8,"mode":0,"inputs":[{"name":"AUDIO_LM","type":"AUDIO_MODEL","link":10},{"name":"VOCODER","type":"AUDIO_MODEL","link":11}],"outputs":[{"name":"AUDIO","type":"AUDIO","links":[12],"shape":3}],"properties":{"Node name for S&R":"LTXGenerateAudio"},"widgets_values":["atmospheric cinematic background music, calm, immersive",8]},{"id":12,"type":"PreviewVideoAudio","pos":[771,530],"size":[280,300],"flags":{},"order":9,"mode":0,"inputs":[{"name":"video","type":"IMAGE","link":8},{"name":"audio","type":"AUDIO","link":12}],"outputs":[],"properties":{"Node name for S&R":"PreviewVideoAudio"},"widgets_values":["ltx_av_lowvram","video/mp4",1]}],"links":[[2,1,0,5,0],[3,2,0,3,0],[4,1,1,5,1],[5,3,0,5,2],[6,4,0,5,3],[7,5,0,6,1],[8,6,0,12,0],[10,9,0,11,0],[11,10,0,11,1],[12,11,0,12,1]],"groups":[],"config":{},"extra":{},"version":0.4}
  

  ---

  六、必开的显存优化开关

  在 LTXSampler 里：
  • Offload：enable（必须开）
  • Dtype：FP8
  • 不要开任何增强、高清修复

• 一、原生支持 FP8 加速（推荐，速度最快）

  这类显卡硬件级支持 FP8 计算，加载与推理最稳、速度最快。

  消费级 / 工作站

  • RTX 5090 / 5080 / 5070 Ti（32GB/24GB/16GB）
  • RTX 4090 / 4090 Ti / 4080 Ti / 4080 Super（24GB/24GB/20GB/16GB）
  • RTX A6000 / A6000 Ada / A5000 / A4500（48GB/48GB/24GB/24GB）
  • RTX 4070 Ti Super / 4070 Ti / 4070 Super（16GB/12GB/12GB）

  数据中心 / 专业卡

  • H100 / H200 / L40S / L40
  • A100 80GB / A100 40GB（部分驱动 / 框架支持）

  最低显存门槛

  • 原生 FP8 流畅跑：≥ 24GB 显存（4090/5090/A6000 等）
  • 优化后可跑：≥ 16GB 显存（4070 Ti Super/4080 Super 等，需开 offload + 低分辨率 / 低帧数）

  ---

  二、兼容运行（无原生 FP8，但可强制加载）

  30 系列及更早显卡不支持硬件 FP8，但可通过软件模拟 / 精度转换运行，速度慢、显存占用高、易报错。

  可兼容列表

  • RTX 3090 / 3090 Ti / 3080 Ti / 3080（24GB/24GB/12GB/10GB）
  • RTX A10 / A30 / A40（24GB/24GB/48GB）
  • RTX 2080 Ti / Titan RTX（11GB/24GB，仅极限精简版）

  兼容条件

  • 显存 ≥ 12GB（仅 8 帧 / 640×384 极限精简）
  • 必须开启：--fp8-upcast + --lowvram + --cpu-offload
  • 只能用 distilled-fp8 蒸馏版，不建议用 dev-fp8 完整版

  ---

  三、不支持（无法运行）

  • RTX 2070 及以下（显存＜10GB，无 FP8 兼容）
  • AMD 显卡（ROCm 对 LTX-2 FP8 支持极差，几乎无法跑）
  • Intel Arc 显卡（驱动与框架暂不支持 LTX-2 FP8 推理）
  • 笔记本独显（RTX 4060 及以下，8GB 显存）：仅能跑 distilled-fp4，无法跑 dev-fp8

  ---

  四、显卡选择速查表（按体验排序）

  表格

  显卡	显存	FP8 支持	跑 dev-fp8 体验
  RTX 5090 / A6000	32GB/48GB	✅ 原生	流畅，1024×576 / 16 帧
  RTX 4090 / 4090 Ti	24GB	✅ 原生	流畅，768×512 / 16 帧
  RTX 4080 Ti / 4080 Super	20GB/16GB	✅ 原生	良好，640×384 / 8–12 帧
  RTX 4070 Ti Super	16GB	✅ 原生	一般，640×384 / 8 帧
  RTX 3090 / 3090 Ti	24GB	❌ 软件模拟	慢，易 OOM，仅低分辨率
  RTX 3080 Ti / 3080	12GB/10GB	❌ 软件模拟	极慢，仅极限精简

  ---

  五、快速判断你的显卡

  1. 看型号：40/50 系列 Ada 架构 → 优先选
  2. 看显存：≥ 24GB → 原生流畅；16GB → 优化可跑；＜12GB → 不建议
  3. 看架构：Ampere（30 系）及更早 → 仅兼容，体验差