LTX 2.5 GGUF ComfyUI | Cinematic Image-to-Video
LTX-2.5 GGUF turns images into cinematic videos with synced audio.
AI model workflow
LTX ComfyUI Workflows
Create video and synchronized audio with ready-to-run LTX ComfyUI workflows for text-to-video, image-to-video, keyframes, multishot scenes, lip sync, camera control, and video editing. Run them on RunComfy Cloud, download workflow JSON, or deploy supported workflows as APIs.
36 workflows
Featured workflow
LTX 2.5 ComfyUI Image to Video | Realistic Motion & Native Audio
The workflows
36 LTX ComfyUI Workflows
Each LTX ComfyUI example below is a real video workflow for LTX Video, LTX-2, LTX 2.3, LTX 2.5, LTX 2.5 GGUF, LTX Director, Sulphur, Prompt Relay, IC-LoRA, ID-LoRA, ControlNet, first-last-frame, lip sync, image-to-video, text-to-video, and video editing. Run it on RunComfy Cloud, download the workflow JSON, or deploy supported workflows as APIs.
LTX 2.5 First & Last Frame to Video | Smooth Transitions
First-to-last frame animation delivers controlled cinematic motion with synced audio.
LTX 2.3 Director 2.0 | Pro Timeline Video Editor
LTX-2.3 Director delivers longer, sharper videos with total timeline control.
LTX 2.5 ComfyUI Text to Video | Native Multishot & 4K HDR
LTX-2.5 turns prompts into sharp cinematic videos with synced audio.
LTX 2.5 ComfyUI Image to Video | Realistic Motion & Native Audio
LTX-2.5 turns one frame into sharp video with synced audio.
LTX 2.3 Talking and Singing Lip Sync Video Maker
Turn faces into singing, talking videos with perfect lip sync magic.
LTX 2.3 - Video Face Swap | Seamless Face Replace
Swap faces in video with natural results and zero flicker.
LTX 2.3 Image to Video | Cinematic Motion Creator
Turn images into realistic, cinematic videos with smooth, consistent motion.
LTX 2.3 Close-Up Shots - Cinematic Image-to-Video
Turn portraits into sharp, cinematic close-up videos with native audio.
LTX 2.3 Inpaint | Precise Video Edit LoRA
Edit video areas fast with ultra-precise control and flawless continuity.
LTX 2.3 Sulphur image to video workflow | Cinematic Generator
Turns still images into cinematic motion scenes fast and smooth.
LTX 2.3 Sulphur 2 text to video workflow | Cinematic Generator
Turn text into cinematic character videos with synced motion fast.
LTX 2.3 Movie Builder Workflow | AI Film Generator
Turn prompts into full cinematic stories with total creative control.
LTX 2.3 Outpainting | Seamless Video Frame Expansion
Expand video scenes smoothly with consistent motion and perspective.
LTX 2.3 Multi-View Videos | Multi-Camera Generator
Turns single video into consistent multi-camera scene shots fast.
LTX 2.3 IC-LoRA | Motion Track AI Video Generator
V2V Motion-guided video tool for ultra-precise, style-free AI creation
LTX 2.3 IC-LoRA Camera Control | Video Motion Tool
Recreates cinematic camera motion with smooth, controllable precision.
LTX 2.3 Edit Anything | Smart Video Editor
Edit videos fast. Keep motion clear. Total visual control.
LTX 2.3 Director | Cinematic AI Video Creator
Turn prompts into full cinematic videos with timeline control.
LTX Video | Image+Text to Video
Generates videos from image+text prompts.
LTX 2.3 MSR | Multi-Subject Video Generator
Keeps every character consistent across complex video scenes.
LTX 2.3 Sulphur 2 Prompt Relay Workflow | Cinematic Video Maker
Turns still images into cinematic motion-controlled videos instantly.
LTX 2.3 ID-LoRA | Talking Avatar Generator
Creates true-to-life talking avatars with synced voice and visuals.
LTX 2.3 First Last Frame | Seamless Video Generator
Transforms keyframes into ultra-smooth, realistic video transitions fast.
LTX-2.3 ICLoRA LipDub Workflow | Lip Sync Video Generator
Turn any video into a perfect lip-synced talking masterpiece.
LTX-2 ComfyUI | Real-Time Video Generator
Create real-time videos instantly, faster than any other generator.
LTX 2.3 Dual Character Lip Sync LoRA | Talking Scene Builder
Build lifelike dual-character conversations with synced speech and expressions.
ComfyUI Video Ads With Krea 2 and LTX 2.3 | Product Ads Generator
Krea 2 and LTX 2.3 turn products into cinematic ads fast.
LTX-2 First Last Frame | Key Frames Video Generator
Turn still frames into seamless video and sound transitions fast.
LTX 2.3 Sulphur T2V workflow | Cinematic Text-to-Video Generator
Turns your text into cinematic, expressive story-driven video scenes instantly.
LTX-2 ControlNet | Precision Video Generator
Sharp control, perfect sync, super clear AI video creation.
LTX 2.3 ComfyUI | High-Quality AI Video Generator
Turn text prompts into cinematic AI videos fast and beautifully.
32 of 36 workflows shown
About LTX
What is a LTX ComfyUI workflow?
An LTX ComfyUI workflow is an editable video node graph built around Lightricks' LTX model family. It connects an LTX checkpoint with text, image, keyframe, video, and audio inputs, then adds sampling, decoding, enhancement, LoRA, control, or editing nodes as needed. The graph gives you a working starting point for AI video instead of a list of model files to assemble yourself.
Choose LTX 2.5 for the newest text-to-video and image-to-video workflows on this page, including native audio, multishot generation, and first-and-last-frame control. LTX 2.3 provides a broad set of community workflows for timeline direction, camera motion, lip sync, multi-subject video, inpainting, outpainting, and prompt-based editing. Earlier LTX-2 and LTX Video graphs remain useful when their model, LoRA, or ControlNet matches your setup.
On RunComfy, every LTX ComfyUI workflow keeps its real preview, required files, and intended inputs together. Click Run now to open the prepared graph in cloud ComfyUI, or download the LTX workflow JSON for local use. Your local installation still needs the compatible LTX checkpoint, text encoders, VAE, audio components, and custom nodes listed by the workflow.
References and model resources: LTX-2 GitHub, ComfyUI-LTXVideo GitHub, LTX 2.5 documentation, Explore LTX models on Hugging Face, Explore LTX models on Civitai.
Capabilities
What you can create with these LTX workflows
LTX ComfyUI workflows cover video and audio generation from text, images, keyframes, timelines, camera guidance, speech, and existing footage.
Text-to-video with synchronized audio
Turn a written scene description into a cinematic LTX video with generated motion and a synchronized soundtrack.
Image and keyframe animation
Animate a source image or guide a transition with first and last frames while controlling the visual endpoints.
Lip sync and audio-driven video
Create talking or singing portraits whose mouth movement and performance follow the supplied audio.
Multishot, timeline, and camera control
Direct multiple shots, camera movement, pacing, and transitions with timeline and control workflows.
Video editing, inpainting, and outpainting
Change selected content, repair frames, or extend the canvas while preserving motion across the clip.
LTX models on RunComfy
Choose the right LTX model. Direct the complete video workflow.
RunComfy opens LTX workflows in native ComfyUI with their video model, audio path, LoRAs, control weights, and custom nodes prepared. Add the prompt, image, keyframes, footage, or speech the graph expects, then adjust shot timing, camera motion, resolution, guidance, and enhancement nodes. A cloud GPU lets you keep the full multishot or high-resolution pipeline when it is too heavy for local hardware.
Run an LTX workflow nowNative ComfyUI graphs for LTX video and audio
Open text-to-video, image-to-video, keyframe, multishot, lip-sync, and editing pipelines as complete, editable LTX graphs.
LTX checkpoints, LoRAs, and control weights in the cloud
Keep compatible LTX models, identity and motion LoRAs, control weights, audio components, and supporting files with the workflow.
Custom nodes for directing and editing video
Use timeline, camera-control, lip-sync, inpainting, outpainting, enhancement, and upscaling nodes without reconstructing the local environment.
Cloud GPUs for multishot and high-resolution LTX
Match GPU memory to model size, video length, resolution, audio, and enhancement steps instead of simplifying the graph to fit locally.
Applications
What people use these LTX workflows for
Choose by the video you need to deliver, from short films and product ads to talking avatars, controlled camera shots, face swaps, and expanded frames.
AI Short Films
Build connected scenes with planned shots, visual continuity, and transitions for narrative video.
Product Video Ads
Turn product imagery into polished promotional clips with camera movement and cinematic lighting.
Talking and Singing Avatars
Animate a portrait or character from dialogue or vocals with synchronized facial performance.
Cinematic Camera Motion
Transfer or specify camera movement for controlled pans, pushes, or tracking shots.
Video Face Swap
Replace a face in moving footage while preserving expression, lighting, and temporal consistency.
Video Frame Expansion
Outpaint beyond the original video frame to create wider compositions without abrupt edge changes.
Frequently asked questions
LTX workflow FAQ
Answers for choosing LTX 2.5 or LTX 2.3 workflows, using synced audio and lip sync, controlling first and last frames, checking VRAM needs, and downloading JSON.
What is the best LTX ComfyUI workflow to start with?
Start with an LTX 2.5 image-to-video workflow if you have a reference image, because it gives the model a clear first frame and can generate synchronized audio. Use text-to-video when you want to create the whole scene from a prompt, and use first-and-last-frame workflows when the start and end compositions both matter.
Does LTX 2.5 create video with audio in ComfyUI?
Yes, when you choose an LTX 2.5 workflow that includes the native audio path. The model can generate synchronized video and audio from text, image, or video inputs, but older LTX workflows may be video-only. Check whether the graph asks for audio components before assuming sound will be produced.
Should I use LTX 2.5 or LTX 2.3?
Use LTX 2.5 for newer image-to-video, text-to-video, first-and-last-frame, native audio, and multishot workflows. Use LTX 2.3 when you need an established audio-driven, lip-sync, LoRA, Director, timeline, or camera-control graph that was built for that version. Do not swap the checkpoint unless the workflow says the version is compatible.
Can LTX do lip sync from an image and audio?
Yes. Choose an LTX audio-to-video or lip-sync workflow when you want a portrait or character to speak from an audio file. Single-speaker workflows are better for one talking face, while dual-character workflows are designed for dialogue scenes. Use a clean face image and clear speech for more stable mouth movement.
How much VRAM do I need for LTX 2.5 in ComfyUI?
VRAM depends on model format, resolution, frame count, audio generation, upscaling, and whether the graph uses quantized weights. Community low-VRAM workflows often rely on FP8 or GGUF variants, but they may run slower or require lower resolution. Use Run now when you want the complete cloud workflow without fitting every model file into a local GPU.
How do I control the first and last frame in an LTX video?
Use an LTX first-last-frame workflow when the clip needs to begin and end on specific images. These workflows are better than a plain prompt for planned transitions, product reveals, and shot continuity. If the motion drifts, add clearer action text or use a timeline/keyframe workflow that gives the model more guidance.
Where can I download an LTX ComfyUI workflow JSON?
Click Download JSON on an LTX workflow card or detail page and open the file in ComfyUI. Local use still requires the matching LTX checkpoint, encoders, VAE, audio components, LoRAs, and custom nodes. Click Run now when you want the prepared cloud graph and run panel instead of completing local setup first.
Related resources
More ways to use LTX
Train an LTX LoRA with the supported LTX-2, LTX-2.3, or LTX-2.5 base, or use LTX 2.5 directly from RunComfy Models.
Train an LTX-2 LoRA
Train a custom visual or motion concept against the original supported LTX-2 base.
Open LTX-2 TrainerTrain an LTX-2.3 LoRA
Create an identity, style, or motion LoRA for the LTX-2.3 workflows on this page.
Open LTX-2.3 TrainerTrain an LTX-2.5 LoRA
Train against the newest supported LTX-2.5 base for compatible video generation workflows.
Open LTX-2.5 TrainerUse LTX 2.5 models directly
Open the LTX 2.5 Models page when you want direct video generation instead of starting from a workflow graph.
Browse LTX 2.5 Models