{"id":8127,"date":"2026-05-12T00:45:00","date_gmt":"2026-05-11T16:45:00","guid":{"rendered":"https:\/\/infernews.com\/blog\/?page_id=8127"},"modified":"2026-07-09T19:28:00","modified_gmt":"2026-07-09T11:28:00","slug":"ltx","status":"publish","type":"page","link":"https:\/\/infernews.com\/blog\/ltx\/","title":{"rendered":"LTX 2.3"},"content":{"rendered":"\n<h2 class=\"wp-block-heading\">Setup<\/h2>\n\n\n<figure class=\"wp-block-embed-youtube wp-block-embed is-type-video is-provider-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"lyte-wrapper\" title=\"LTX-2.3 Quickstart in ComfyUI: Install, Models, and Your First Workflow\" style=\"width:853px;max-width:100%;margin:5px auto;\"><div class=\"lyMe\" id=\"WYL__YGc2Tf8lcY\" itemprop=\"video\" itemscope itemtype=\"https:\/\/schema.org\/VideoObject\"><div><meta itemprop=\"thumbnailUrl\" content=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2F_YGc2Tf8lcY%2Fhqdefault.jpg\" \/><meta itemprop=\"embedURL\" content=\"https:\/\/www.youtube.com\/embed\/_YGc2Tf8lcY\" \/><meta itemprop=\"duration\" content=\"PT14M17S\" \/><meta itemprop=\"uploadDate\" content=\"2026-07-08T13:31:37Z\" \/><\/div><div id=\"lyte__YGc2Tf8lcY\" data-src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2F_YGc2Tf8lcY%2Fhqdefault.jpg\" class=\"pL\"><div class=\"tC\"><div class=\"tT\" itemprop=\"name\">LTX-2.3 Quickstart in ComfyUI: Install, Models, and Your First Workflow<\/div><\/div><div class=\"play\"><\/div><div class=\"ctrl\"><div class=\"Lctrl\"><\/div><div class=\"Rctrl\"><\/div><\/div><\/div><noscript><a href=\"https:\/\/youtu.be\/_YGc2Tf8lcY\" rel=\"nofollow noopener\" target=\"_blank\"><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2F_YGc2Tf8lcY%2F0.jpg\" alt=\"LTX-2.3 Quickstart in ComfyUI: Install, Models, and Your First Workflow\" width=\"853\" height=\"460\" \/><br \/>Watch this video on YouTube<\/a><\/noscript><meta itemprop=\"description\" content=\"LTX Quickstart, Part 1: ComfyUI Install and First Workflow In part one of the LTX quickstart, we walk through installing the LTX-2.3 custom node for ComfyUI from source. You&#039;ll see how to git clone the repository, run UV sync with locked dependencies, activate the environment, and load an example workflow from the 2.3 folder. The session also covers where to find LTX-2.3 checkpoints and LoRAs on HuggingFace, how to download the Gemma 3 text encoder, where each model file belongs inside your ComfyUI directory, VRAM requirements for local generation, and how to fix common workflow issues like the bypass toggle for image-to-video. This is for developers, ComfyUI users, and technical artists comfortable with Python, git, and the command line who want to run LTX-2.3 locally on their own machine. 00:00 - 00:42 Intro: LTX Quickstart in ComfyUI 00:42 - 01:07 Who This Video Is For 01:07 - 01:34 Getting ComfyUI Desktop Installed 01:34 - 02:24 Two Install Paths: LTX Desktop vs Git Method 02:24 - 03:25 Git Clone the LTX Video Custom Node 03:25 - 03:51 Why UV Sync Frozen Matters 03:51 - 04:47 Terminal Walkthrough on Windows 04:47 - 05:39 Activating the Environment and Modifying the Code 05:39 - 06:25 Example Workflows in the 2.3 Directory 06:25 - 07:25 Loading the Two-Stage Distilled Workflow 07:25 - 08:17 Downloading Models from HuggingFace and Gemma 3 08:17 - 09:05 Where Checkpoints and LoRAs Belong 09:05 - 09:57 Hardware and VRAM Requirements 09:57 - 10:37 The Rig Running This Tutorial 10:37 - 11:31 Opening the Workflow and Fixing the Bypass Toggle 11:31 - 13:00 Pointing Model Paths for LTX-2.3, LoRA, and Gemma 13:00 - 13:23 Loading the Base Image for Image-to-Video 13:23 - 14:16 Recap and What&#039;s Next in Part 2 \ud83d\udd37 Join our Discord Community: https:\/\/discord.com\/invite\/ltxplatform \ud83d\udd37 Follow us on X: https:\/\/x.com\/ltx_io \ud83d\udd37 Follow us on Instagram: https:\/\/www.instagram.com\/ltx.io\/ About LTX LTX builds open world models: AI that understands and generates physical reality across video, audio, and multimodal inputs. Open, efficient, and yours to own. Run it on your own hardware and fine-tune it on your own IP, with 18M+ downloads and no vendor lock-in. \ud83d\udd14 Subscribe for more LTX model releases, open-source tools, training guides, and developer resources.\"><\/div><\/div><div class=\"lL\" style=\"max-width:100%;width:853px;margin:5px auto;\"><\/div><figcaption><\/figcaption><\/figure>\n\n<figure class=\"wp-block-embed-youtube wp-block-embed is-type-video is-provider-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"lyte-wrapper\" title=\"LTX-2.3 Quickstart Part 2: Prompt Structure, Rendering, and Debugging in ComfyUI\" style=\"width:853px;max-width:100%;margin:5px auto;\"><div class=\"lyMe\" id=\"WYL_6Ijwv2SH11I\" itemprop=\"video\" itemscope itemtype=\"https:\/\/schema.org\/VideoObject\"><div><meta itemprop=\"thumbnailUrl\" content=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2F6Ijwv2SH11I%2Fhqdefault.jpg\" \/><meta itemprop=\"embedURL\" content=\"https:\/\/www.youtube.com\/embed\/6Ijwv2SH11I\" \/><meta itemprop=\"duration\" content=\"PT13M45S\" \/><meta itemprop=\"uploadDate\" content=\"2026-07-08T13:31:42Z\" \/><\/div><div id=\"lyte_6Ijwv2SH11I\" data-src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2F6Ijwv2SH11I%2Fhqdefault.jpg\" class=\"pL\"><div class=\"tC\"><div class=\"tT\" itemprop=\"name\">LTX-2.3 Quickstart Part 2: Prompt Structure, Rendering, and Debugging in ComfyUI<\/div><\/div><div class=\"play\"><\/div><div class=\"ctrl\"><div class=\"Lctrl\"><\/div><div class=\"Rctrl\"><\/div><\/div><\/div><noscript><a href=\"https:\/\/youtu.be\/6Ijwv2SH11I\" rel=\"nofollow noopener\" target=\"_blank\"><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2F6Ijwv2SH11I%2F0.jpg\" alt=\"LTX-2.3 Quickstart Part 2: Prompt Structure, Rendering, and Debugging in ComfyUI\" width=\"853\" height=\"460\" \/><br \/>Watch this video on YouTube<\/a><\/noscript><meta itemprop=\"description\" content=\"LTX Quickstart, Part 2: Prompting, Rendering, and Debugging In part two of the LTX quickstart, we finish the image-to-video workflow you set up in part one. You&#039;ll see how LTX prompts are structured (subject, action, setting, camera move, audio cues), how to use an LLM with the LTX prompting guide to expand a simple idea into a production prompt, and how to configure FPS, frame count, and video size before hitting run. The session also covers reading the workflow graph left to right, how the two-stage distilled workflow runs low-resolution generation then upscales to 1920x1088, why the first render is slower than the ones that follow, and how to debug common ComfyUI issues using the console, node reloads, and UV sync frozen for dependency mismatches. This is for developers, ComfyUI users, and technical artists who finished part one and want to generate their first LTX-2.3 video locally with synchronized sound effects. 00:00 - 00:24 Intro: Welcome Back to Part 2 00:24 - 01:19 Opening the Image-to-Video Workflow 01:19 - 02:02 Reading the Workflow Graph Left to Right 02:02 - 02:46 LTX Prompt Structure: Subject, Action, Setting, Camera, Audio 02:46 - 03:27 Using the LTX Prompting Guide with an LLM 03:27 - 04:36 Building the Pinball Prompt 04:36 - 04:55 Sound Effect Cues Inside the Prompt 04:55 - 05:37 FPS, Frame Count, and Video Size 05:37 - 06:00 Latent Video Node Settings 06:00 - 06:30 Hitting Run and Hardware Expectations 06:30 - 07:00 First Run Load Times 07:00 - 07:51 Two-Stage Workflow: Low Res Then Upscale 07:51 - 08:25 Reviewing the Final Video with Synced Audio 08:25 - 08:47 Gotcha: Restart ComfyUI After New Nodes 08:47 - 09:30 Debugging with the ComfyUI Console 09:30 - 10:10 Using an LLM to Read Console Errors 10:10 - 10:51 Dependency Mismatches and UV Sync Frozen 10:51 - 11:22 Where to Go Next: Prompting and Image-to-Video Guides 11:22 - 12:15 LTX Trainer and the GitHub Docs 12:15 - 12:38 Kudos to the LTX Documentation 12:38 - 13:22 Join the LTX Discord Community 13:22 - 13:44 Final Thoughts &amp; Thanks \ud83d\udd37 Join our Discord Community: https:\/\/discord.com\/invite\/ltxplatform \ud83d\udd37 Follow us on X: https:\/\/x.com\/ltx_io \ud83d\udd37 Follow us on Instagram: https:\/\/www.instagram.com\/ltx.io\/ About LTX LTX builds open world models: AI that understands and generates physical reality across video, audio, and multimodal inputs. Open, efficient, and yours to own. Run it on your own hardware and fine-tune it on your own IP, with 18M+ downloads and no vendor lock-in. \ud83d\udd14 Subscribe for more LTX model releases, open-source tools, training guides, and developer resources.\"><\/div><\/div><div class=\"lL\" style=\"max-width:100%;width:853px;margin:5px auto;\"><\/div><figcaption><\/figcaption><\/figure>\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<h2 class=\"wp-block-heading\">All in one<\/h2>\n\n\n<figure class=\"wp-block-embed-youtube wp-block-embed is-type-video is-provider-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"lyte-wrapper\" title=\"How to Enhance Videos with LTX 2.3 | Remove Watermarks &amp;amp; Text, Upscale &amp;amp; Outpaint\" style=\"width:853px;max-width:100%;margin:5px auto;\"><div class=\"lyMe\" id=\"WYL_bAN7XdRa-20\" itemprop=\"video\" itemscope itemtype=\"https:\/\/schema.org\/VideoObject\"><div><meta itemprop=\"thumbnailUrl\" content=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FbAN7XdRa-20%2Fhqdefault.jpg\" \/><meta itemprop=\"embedURL\" content=\"https:\/\/www.youtube.com\/embed\/bAN7XdRa-20\" \/><meta itemprop=\"duration\" content=\"PT4M26S\" \/><meta itemprop=\"uploadDate\" content=\"2026-07-06T12:15:17Z\" \/><\/div><meta itemprop=\"accessibilityFeature\" content=\"captions\" \/><div id=\"lyte_bAN7XdRa-20\" data-src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FbAN7XdRa-20%2Fhqdefault.jpg\" class=\"pL\"><div class=\"tC\"><div class=\"tT\" itemprop=\"name\">How to Enhance Videos with LTX 2.3 | Remove Watermarks &amp; Text, Upscale &amp; Outpaint<\/div><\/div><div class=\"play\"><\/div><div class=\"ctrl\"><div class=\"Lctrl\"><\/div><div class=\"Rctrl\"><\/div><\/div><\/div><noscript><a href=\"https:\/\/youtu.be\/bAN7XdRa-20\" rel=\"nofollow noopener\" target=\"_blank\"><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FbAN7XdRa-20%2F0.jpg\" alt=\"How to Enhance Videos with LTX 2.3 | Remove Watermarks &amp;amp; Text, Upscale &amp;amp; Outpaint\" width=\"853\" height=\"460\" \/><br \/>Watch this video on YouTube<\/a><\/noscript><meta itemprop=\"description\" content=\"Looking for an easy way to remove watermarks, subtitles, restore blurry videos, or even convert vertical videos into horizontal ones\u2014all inside a single ComfyUI workflow? In this tutorial, I&#039;ll show you my latest LTX 2.3 All-in-One Video Enhancement Workflow, designed to handle multiple video editing tasks with a single sampler while keeping VRAM usage as low as possible. In this video, you&#039;ll learn how to: \u2022 Remove watermarks, logos, icons, and subtitles from videos \u2022 Restore and upscale low-quality footage to HD \u2022 Expand videos horizontally or vertically with AI Outpainting \u2022 Configure the workflow for the best balance between quality and performance Unlike the traditional LTX 2.3 pipeline, this workflow bypasses the standard downscale\/upscale process to preserve as much detail as possible. Although it takes a little longer to render, the improvement in visual quality is well worth it. \ud83d\udcac If you have any questions or ideas for future workflows, feel free to leave a comment below. I always enjoy hearing your feedback. If you find this video helpful, don&#039;t forget to Like, Subscribe, and turn on notifications so you won&#039;t miss future ComfyUI workflows, AI video tutorials, and workflow updates. \u2728 All workflows and resources are shared for free. If you\u2019d like to support the channel, you can buy me a coffee \u2764\ufe0f: https:\/\/ko-fi.com\/fututek -------------------------------------------------- \u2728Run on Cloud GPU: https:\/\/runpod.io?ref=99cuxxw1 -------------------------------------------------- \u2728Workflow and resources: https:\/\/drive.google.com\/drive\/folders\/1Eby6vrdr3dyINW7AaigGy6WxTMw3AAj9?usp=sharing - LTX2.3 Model (Place in \/ComfyUI\/models\/diffusion_models) https:\/\/huggingface.co\/vantagewithai\/LTX-2.3-GGUF\/blob\/main\/dev\/ltx-2-3-22b-dev-Q4_K_M.gguf - Text Encoder (Place in \/ComfyUI\/models\/text_encoders) https:\/\/huggingface.co\/Comfy-Org\/ltx-2\/resolve\/main\/split_files\/text_encoders\/gemma_3_12B_it_fp8_scaled.safetensors https:\/\/huggingface.co\/Kijai\/LTX2.3_comfy\/blob\/main\/text_encoders\/ltx-2.3_text_projection_bf16.safetensors - VAE (Place in \/ComfyUI\/models\/vae) https:\/\/huggingface.co\/Kijai\/LTX2.3_comfy\/blob\/main\/vae\/LTX23_audio_vae_bf16.safetensors https:\/\/huggingface.co\/Kijai\/LTX2.3_comfy\/blob\/main\/vae\/LTX23_video_vae_bf16.safetensors - Distilled Loras (Place in \/ComfyUI\/models\/loras) https:\/\/huggingface.co\/Kijai\/LTX2.3_comfy\/blob\/main\/loras\/ltx-2.3-22b-distilled-lora-dynamic_fro09_avg_rank_105_bf16.safetensors - Enhancement Loras (Place in \/ComfyUI\/models\/loras) https:\/\/huggingface.co\/joyfox\/LTX2.3-ICEdit-Insight\/resolve\/main\/ltx2.3-ic-subtitles-remove-general.safetensors?download=true https:\/\/huggingface.co\/joyfox\/LTX2.3-ICEdit-Insight\/resolve\/main\/ltx2.3-ic-video-upscale-general.safetensors?download=true https:\/\/huggingface.co\/joyfox\/LTX2.3-ICEdit-Insight\/resolve\/main\/ltx2.3-ic-watermark-remove-general.safetensors?download=true https:\/\/huggingface.co\/joyfox\/LTX2.3-ICEdit-Insight\/resolve\/main\/ltx2.3-video-restoration-general.safetensors?download=true https:\/\/huggingface.co\/oumoumad\/LTX-2.3-22b-IC-LoRA-Outpaint\/resolve\/main\/ltx-2.3-22b-ic-lora-outpaint.safetensors?download=true https:\/\/huggingface.co\/Lightricks\/LTX-2-19b-IC-LoRA-Detailer\/resolve\/main\/ltx-2-19b-ic-lora-detailer.safetensors\"><\/div><\/div><div class=\"lL\" style=\"max-width:100%;width:853px;margin:5px auto;\"><\/div><figcaption><\/figcaption><\/figure>\n\n<figure class=\"wp-block-embed-youtube wp-block-embed is-type-video is-provider-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"lyte-wrapper\" title=\"LTX 2.3 Ingredients IC-LoRA vs MSR &mdash; Which Reference LoRA Wins in ComfyUI?\" style=\"width:853px;max-width:100%;margin:5px auto;\"><div class=\"lyMe\" id=\"WYL_jNWZL2azMjs\" itemprop=\"video\" itemscope itemtype=\"https:\/\/schema.org\/VideoObject\"><div><meta itemprop=\"thumbnailUrl\" content=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FjNWZL2azMjs%2Fhqdefault.jpg\" \/><meta itemprop=\"embedURL\" content=\"https:\/\/www.youtube.com\/embed\/jNWZL2azMjs\" \/><meta itemprop=\"duration\" content=\"PT12M56S\" \/><meta itemprop=\"uploadDate\" content=\"2026-07-03T13:52:16Z\" \/><\/div><div id=\"lyte_jNWZL2azMjs\" data-src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FjNWZL2azMjs%2Fhqdefault.jpg\" class=\"pL\"><div class=\"tC\"><div class=\"tT\" itemprop=\"name\">LTX 2.3 Ingredients IC-LoRA vs MSR \u2014 Which Reference LoRA Wins in ComfyUI?<\/div><\/div><div class=\"play\"><\/div><div class=\"ctrl\"><div class=\"Lctrl\"><\/div><div class=\"Rctrl\"><\/div><\/div><\/div><noscript><a href=\"https:\/\/youtu.be\/jNWZL2azMjs\" rel=\"nofollow noopener\" target=\"_blank\"><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FjNWZL2azMjs%2F0.jpg\" alt=\"LTX 2.3 Ingredients IC-LoRA vs MSR &mdash; Which Reference LoRA Wins in ComfyUI?\" width=\"853\" height=\"460\" \/><br \/>Watch this video on YouTube<\/a><\/noscript><meta itemprop=\"description\" content=\"Workflow- https:\/\/buymeacoffee.com\/ai_jigyasa\/e\/553772 Paid Workflow with local Asset (reference sheet Generator)- https:\/\/buymeacoffee.com\/ai_jigyasa\/e\/554470 In this video, I&#039;ll show you how to generate AI videos directly from a reference sheet using LTX 2.3 and IC-LoRA inside ComfyUI \u2014 no character setup, no separate image-to-video step, just drop in your reference sheet and let the model do the work. This workflow tests and compares the Ingredients IC-LoRA by Lightricks and the MSR (Multiple Subject Reference) IC-LoRA by LiconStudio side by side, so you can see exactly which one holds character consistency better for your use case. Featuring IC-LoRA Ingredients: https:\/\/huggingface.co\/Lightricks\/LTX-2.3-22b-IC-LoRA-Ingredients Featuring IC-LoRA MSR: https:\/\/huggingface.co\/LiconStudio\/LTX-2.3-Multiple-Subject-Reference What you&#039;ll learn: LTX 2.3 Audio-Video Model IC-LoRA Reference Editing ComfyUI Workflow &amp; Resource Monitoring LoRA Comparison (Ingredients vs. Lykon) Refinement &amp; Enhancer LoRAs Technical Generation Settings Character Sheet Prompting Automated Prompt Generation Model links Models (dev) BF16: https:\/\/huggingface.co\/Lightricks\/LTX-2.3\/resolve\/main\/ltx-2.3-22b-dev.safetensors?download=true FP8: https:\/\/huggingface.co\/Kijai\/LTX2.3_comfy\/resolve\/main\/diffusion_models\/ltx-2.3-22b-dev_transformer_only_fp8_scaled.safetensors?download=true GGUF: https:\/\/huggingface.co\/QuantStack\/LTX-2.3-GGUF\/tree\/main Models (distilled) BF16: https:\/\/huggingface.co\/Lightricks\/LTX-2.3\/resolve\/main\/ltx-2.3-22b-distilled-1.1.safetensors?download=true FP8: https:\/\/huggingface.co\/Kijai\/LTX2.3_comfy\/resolve\/main\/diffusion_models\/ltx-2.3-22b-distilled_transformer_only_fp8_scaled.safetensors?download=true GGUF: https:\/\/huggingface.co\/QuantStack\/LTX-2.3-GGUF\/tree\/main text_encoders https:\/\/huggingface.co\/Comfy-Org\/ltx-2\/resolve\/main\/split_files\/text_encoders\/gemma_3_12B_it.safetensors?download=true https:\/\/huggingface.co\/Kijai\/LTX2.3_comfy\/resolve\/main\/text_encoders\/ltx-2.3_text_projection_bf16.safetensors?download=true vae https:\/\/huggingface.co\/Kijai\/LTX2.3_comfy\/resolve\/main\/vae\/LTX23_video_vae_bf16.safetensors?download=true https:\/\/huggingface.co\/Kijai\/LTX2.3_comfy\/resolve\/main\/vae\/LTX23_audio_vae_bf16.safetensors?download=true loras (distilled) https:\/\/huggingface.co\/Lightricks\/LTX-2.3\/resolve\/main\/ltx-2.3-22b-distilled-lora-384-1.1.safetensors?download=true IC-LoRAs \ud83d\udd39 Ingredients (by Lightricks) \u2014 Reference sheet \u2192 video consistency More info: https:\/\/huggingface.co\/Lightricks\/LTX-2.3-22b-IC-LoRA-Ingredients https:\/\/huggingface.co\/Lightricks\/LTX-2.3-22b-IC-LoRA-Ingredients\/resolve\/main\/ltx-2.3-22b-ic-lora-ingredients-0.9.safetensors?download=true \ud83d\udd39 MSR \u2014 Multiple Subject Reference (by LiconStudio) \u2014 Multi-character reference consistency More info: https:\/\/huggingface.co\/LiconStudio\/LTX-2.3-Multiple-Subject-Reference V1: https:\/\/huggingface.co\/LiconStudio\/LTX-2.3-Multiple-Subject-Reference\/resolve\/main\/LTX-2.3-Licon-MSR-V1.safetensors?download=true Test Version: https:\/\/huggingface.co\/LiconStudio\/LTX-2.3-Multiple-Subject-Reference\/resolve\/main\/LTX2.3-Licon-MSR-test_version.safetensors?download=true \ud83d\udd39 Union Control (by Lightricks) \u2014 Multi-control reference https:\/\/huggingface.co\/Lightricks\/LTX-2.3-22b-IC-LoRA-Union-Control\/resolve\/main\/ltx-2.3-22b-ic-lora-union-control-ref0.5.safetensors?download=true latent_upscale_models https:\/\/huggingface.co\/Lightricks\/LTX-2.3\/resolve\/main\/ltx-2.3-spatial-upscaler-x2-1.1.safetensors?download=true Vocals Extraction https:\/\/huggingface.co\/Kijai\/MelBandRoFormer_comfy\/resolve\/main\/MelBandRoformer_fp32.safetensors?download=true \ud83d\udc4d Support the Channel If you found this helpful: \ud83d\udc4d Like the video \ud83d\udd14 Subscribe for more AI workflows \ud83d\udcac Comment what you want to see next #LTX23 #ComfyUI #AIVideo #GenerativeAI #ICLoRA #ReferenceSheet #CharacterConsistency #Ref2Edit #LightricksAI #OpenSourceAI #AIContentCreation #ComfyUIWorkflow\"><\/div><\/div><div class=\"lL\" style=\"max-width:100%;width:853px;margin:5px auto;\"><\/div><figcaption><\/figcaption><\/figure>\n\n\n<h2 class=\"wp-block-heading\">LTX Trainer<\/h2>\n\n\n<figure class=\"wp-block-embed-youtube wp-block-embed is-type-video is-provider-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"lyte-wrapper\" title=\"LTX Trainer: Open-Source LoRA Training\" style=\"width:853px;max-width:100%;margin:5px auto;\"><div class=\"lyMe\" id=\"WYL_7N2kTJQCgEk\" itemprop=\"video\" itemscope itemtype=\"https:\/\/schema.org\/VideoObject\"><div><meta itemprop=\"thumbnailUrl\" content=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2F7N2kTJQCgEk%2Fhqdefault.jpg\" \/><meta itemprop=\"embedURL\" content=\"https:\/\/www.youtube.com\/embed\/7N2kTJQCgEk\" \/><meta itemprop=\"duration\" content=\"PT1M21S\" \/><meta itemprop=\"uploadDate\" content=\"2026-06-17T14:41:08Z\" \/><\/div><div id=\"lyte_7N2kTJQCgEk\" data-src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2F7N2kTJQCgEk%2Fhqdefault.jpg\" class=\"pL\"><div class=\"tC\"><div class=\"tT\" itemprop=\"name\">LTX Trainer: Open-Source LoRA Training<\/div><\/div><div class=\"play\"><\/div><div class=\"ctrl\"><div class=\"Lctrl\"><\/div><div class=\"Rctrl\"><\/div><\/div><\/div><noscript><a href=\"https:\/\/youtu.be\/7N2kTJQCgEk\" rel=\"nofollow noopener\" target=\"_blank\"><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2F7N2kTJQCgEk%2F0.jpg\" alt=\"LTX Trainer: Open-Source LoRA Training\" width=\"853\" height=\"460\" \/><br \/>Watch this video on YouTube<\/a><\/noscript><meta itemprop=\"description\" content=\"LTX Trainer is here. Train LTX on your own data. LTX Trainer is the open-source framework for customizing LTX with your own characters, styles, workflows, and IP. Build LoRAs and IC-LoRAs so the model learns your world and reproduces it consistently across projects. Your data and IP never leave your infrastructure. What&#039;s new in this release: \u25cf New conditioning modes out of the box: video, audio, cross-modal, and reference (IC-LoRAs) \u25cf Combine multiple conditioning methods from a single framework \u25cf Agentic setup: use Claude or any LLM to configure your training runs \u25cf Fully open source, maintained by the core LTX team Get started: GitHub: https:\/\/github.com\/Lightricks\/LTX-2\/blob\/main\/packages\/ltx-trainer\/README.md Documentation: https:\/\/docs.ltx.video\/open-source-model\/getting-started\/overview HuggingFace: https:\/\/huggingface.co\/collections\/Lightricks\/ltx-23-creative-lab \ud83d\udd14 Subscribe for updates on LTX model releases, open-source tools, and developer resources.\"><\/div><\/div><div class=\"lL\" style=\"max-width:100%;width:853px;margin:5px auto;\"><\/div><figcaption><\/figcaption><\/figure>\n\n<figure class=\"wp-block-embed-youtube wp-block-embed is-type-video is-provider-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"lyte-wrapper\" title=\"LTX Trainer Explained - Ingredients IC-LoRA The NEW Reference-to-Video Model Explained\" style=\"width:853px;max-width:100%;margin:5px auto;\"><div class=\"lyMe\" id=\"WYL_7nMTLCfohqs\" itemprop=\"video\" itemscope itemtype=\"https:\/\/schema.org\/VideoObject\"><div><meta itemprop=\"thumbnailUrl\" content=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2F7nMTLCfohqs%2Fhqdefault.jpg\" \/><meta itemprop=\"embedURL\" content=\"https:\/\/www.youtube.com\/embed\/7nMTLCfohqs\" \/><meta itemprop=\"duration\" content=\"PT7M5S\" \/><meta itemprop=\"uploadDate\" content=\"2026-06-19T19:08:03Z\" \/><\/div><div id=\"lyte_7nMTLCfohqs\" data-src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2F7nMTLCfohqs%2Fhqdefault.jpg\" class=\"pL\"><div class=\"tC\"><div class=\"tT\" itemprop=\"name\">LTX Trainer Explained - Ingredients IC-LoRA The NEW Reference-to-Video Model Explained<\/div><\/div><div class=\"play\"><\/div><div class=\"ctrl\"><div class=\"Lctrl\"><\/div><div class=\"Rctrl\"><\/div><\/div><\/div><noscript><a href=\"https:\/\/youtu.be\/7nMTLCfohqs\" rel=\"nofollow noopener\" target=\"_blank\"><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2F7nMTLCfohqs%2F0.jpg\" alt=\"LTX Trainer Explained - Ingredients IC-LoRA The NEW Reference-to-Video Model Explained\" width=\"853\" height=\"460\" \/><br \/>Watch this video on YouTube<\/a><\/noscript><meta itemprop=\"description\" content=\"LTX-2 AI video model, LTX Trainer fine-tuning, Ingredients IC-LoRA, consistent AI video characters, open-source AI video generation, and custom LoRA training just got a massive upgrade. Lightricks released LTX 2.3 with a fully overhauled open-source trainer that lets you fine-tune a 22 billion parameter multimodal AI model that generates synchronized video and audio in one single pass \u2014 and you own the weights. In this video we break down how the LTX Trainer works, the difference between LoRA, IC-LoRA, and full fine-tuning, how the Ingredients reference-to-video model solves the AI character consistency problem using a simple reference grid image, and how to automate the entire pipeline using AI agents, ComfyUI, and Gemma 3. This video is for AI filmmakers, creative directors, indie developers, and content creators who want to build hyper-consistent AI video with full control over their models without being locked into expensive cloud platforms. ------------------------------------------------------------------- ETSY Store: https:\/\/www.etsy.com\/shop\/maxonshire Twitter: https:\/\/twitter.com\/MaxonShire ------------------------------------------------------------------- Relevant hashtags: #LTX #LTXTrainer #AIVideo #LoRA #LTX2 #ConsistentAICharacters #OpenSourceAI #AIVideoGeneration #Lightricks #ImageToVideo #TextToVideo #AIAnimation #AI\"><\/div><\/div><div class=\"lL\" style=\"max-width:100%;width:853px;margin:5px auto;\"><\/div><figcaption><\/figcaption><\/figure>\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<h2 class=\"wp-block-heading\">LTX Director<\/h2>\n\n\n<figure class=\"wp-block-embed-youtube wp-block-embed is-type-video is-provider-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"lyte-wrapper\" title=\"LTX Director 2.0 Is INSANE &mdash; Edit Existing Videos in ComfyUI\" style=\"width:853px;max-width:100%;margin:5px auto;\"><div class=\"lyMe\" id=\"WYL_BFE9v9LZpIU\" itemprop=\"video\" itemscope itemtype=\"https:\/\/schema.org\/VideoObject\"><div><meta itemprop=\"thumbnailUrl\" content=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FBFE9v9LZpIU%2Fhqdefault.jpg\" \/><meta itemprop=\"embedURL\" content=\"https:\/\/www.youtube.com\/embed\/BFE9v9LZpIU\" \/><meta itemprop=\"duration\" content=\"PT11M21S\" \/><meta itemprop=\"uploadDate\" content=\"2026-06-26T16:03:57Z\" \/><\/div><div id=\"lyte_BFE9v9LZpIU\" data-src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FBFE9v9LZpIU%2Fhqdefault.jpg\" class=\"pL\"><div class=\"tC\"><div class=\"tT\" itemprop=\"name\">LTX Director 2.0 Is INSANE \u2014 Edit Existing Videos in ComfyUI<\/div><\/div><div class=\"play\"><\/div><div class=\"ctrl\"><div class=\"Lctrl\"><\/div><div class=\"Rctrl\"><\/div><\/div><\/div><noscript><a href=\"https:\/\/youtu.be\/BFE9v9LZpIU\" rel=\"nofollow noopener\" target=\"_blank\"><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FBFE9v9LZpIU%2F0.jpg\" alt=\"LTX Director 2.0 Is INSANE &mdash; Edit Existing Videos in ComfyUI\" width=\"853\" height=\"460\" \/><br \/>Watch this video on YouTube<\/a><\/noscript><meta itemprop=\"description\" content=\"LTX Director 2.0 is finally starting to feel like a real AI video editor inside ComfyUI. In this video, I show how to use the new LTX Director 2.0 workflow to import an existing video, work with the timeline, select a specific fragment, use Retake Mode, and change only the part of the video you want \u2014 without regenerating the entire scene from scratch. We will test several practical examples: * adding lightning behind the characters * replacing coffee cups with bananas * changing outfits into dresses * turning night into day * using local and global prompts correctly * understanding when InPaint is needed * using Edit Anything LoRA for stronger object and clothing changes * preparing the workflow for Full HD generation inside ComfyUI This is not just text-to-video generation anymore. This is the beginning of controlled AI video editing directly inside ComfyUI. Free workflow, model links, and useful materials: https:\/\/www.patreon.com\/AntiMatterComfy\/posts\/ltx-director-2-0-162147683?utm_medium=clipboard_copy&amp;utm_source=copyLink&amp;utm_campaign=postshare_creator&amp;utm_content=join_link In the next video, I will go deeper into IC Video \/ IC-LoRA and how to transfer motion, pose, or style from a reference video into your own character or scene. If you are interested in ComfyUI, LTX 2.3, LTX Director, Flux, local AI video generation, and AI filmmaking \u2014 subscribe to the channel. #LTXDirector #ComfyUI #LTX23 #AIVideo #AIFilmmaking #RetakeMode #LocalAI #ComfyUIWorkflow #AIAnimation #Flux\"><\/div><\/div><div class=\"lL\" style=\"max-width:100%;width:853px;margin:5px auto;\"><\/div><figcaption><\/figcaption><\/figure>\n\n<figure class=\"wp-block-embed-youtube wp-block-embed is-type-video is-provider-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"lyte-wrapper\" title=\"LTX Director + 4 Panel + Global Ref | Local AI Video Workflow\" style=\"width:853px;max-width:100%;margin:5px auto;\"><div class=\"lyMe\" id=\"WYL_rdWmc1BWhKc\" itemprop=\"video\" itemscope itemtype=\"https:\/\/schema.org\/VideoObject\"><div><meta itemprop=\"thumbnailUrl\" content=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FrdWmc1BWhKc%2Fhqdefault.jpg\" \/><meta itemprop=\"embedURL\" content=\"https:\/\/www.youtube.com\/embed\/rdWmc1BWhKc\" \/><meta itemprop=\"duration\" content=\"PT5M\" \/><meta itemprop=\"uploadDate\" content=\"2026-06-07T14:05:05Z\" \/><\/div><meta itemprop=\"accessibilityFeature\" content=\"captions\" \/><div id=\"lyte_rdWmc1BWhKc\" data-src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FrdWmc1BWhKc%2Fhqdefault.jpg\" class=\"pL\"><div class=\"tC\"><div class=\"tT\" itemprop=\"name\">LTX Director + 4 Panel + Global Ref | Local AI Video Workflow<\/div><\/div><div class=\"play\"><\/div><div class=\"ctrl\"><div class=\"Lctrl\"><\/div><div class=\"Rctrl\"><\/div><\/div><\/div><noscript><a href=\"https:\/\/youtu.be\/rdWmc1BWhKc\" rel=\"nofollow noopener\" target=\"_blank\"><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FrdWmc1BWhKc%2F0.jpg\" alt=\"LTX Director + 4 Panel + Global Ref | Local AI Video Workflow\" width=\"853\" height=\"460\" \/><br \/>Watch this video on YouTube<\/a><\/noscript><meta itemprop=\"description\" content=\"Camera lab github link: https:\/\/github.com\/ai2764\/Camera-lab In this video, I show my extended LTX Director setup inside Camera Lab. This feature combines a 4-panel storyboard method, LTX Director-style shot control, and a global reference image system to make local AI video generation easier to direct shot by shot. I also added dependency notes so a coding agent can help pull the project, follow the setup guide, and get it running locally. Credits: Inspired by Tudou\u2019s 4-panel storyboard idea and What Dreams Cost\u2019s LTX Director workflow. #LTXDirector #CameraLab #AIVideoGeneration #LocalAI #ComfyUI #LTXVideo #AIFilmmaking #AIDirector #4PanelStoryboard #GlobalReference #ImageToVideo #LocalVideoGeneration #AIVideoWorkflow #StoryboardToVideo #PromptRelay #DirectorTool #AITools #GenerativeVideo #VideoGenerationTutorial\"><\/div><\/div><div class=\"lL\" style=\"max-width:100%;width:853px;margin:5px auto;\"><\/div><figcaption><\/figcaption><\/figure>\n\n<figure class=\"wp-block-embed-youtube wp-block-embed is-type-video is-provider-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"lyte-wrapper\" title=\"LTX Director 2 + SEED HUNTER workflow | The Dangers of Convenience\/Overhaul Nodes\" style=\"width:853px;max-width:100%;margin:5px auto;\"><div class=\"lyMe\" id=\"WYL_vGxZIS4J0Mg\" itemprop=\"video\" itemscope itemtype=\"https:\/\/schema.org\/VideoObject\"><div><meta itemprop=\"thumbnailUrl\" content=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FvGxZIS4J0Mg%2Fhqdefault.jpg\" \/><meta itemprop=\"embedURL\" content=\"https:\/\/www.youtube.com\/embed\/vGxZIS4J0Mg\" \/><meta itemprop=\"duration\" content=\"PT3M23S\" \/><meta itemprop=\"uploadDate\" content=\"2026-06-21T22:14:45Z\" \/><\/div><div id=\"lyte_vGxZIS4J0Mg\" data-src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FvGxZIS4J0Mg%2Fhqdefault.jpg\" class=\"pL\"><div class=\"tC\"><div class=\"tT\" itemprop=\"name\">LTX Director 2 + SEED HUNTER workflow | The Dangers of Convenience\/Overhaul Nodes<\/div><\/div><div class=\"play\"><\/div><div class=\"ctrl\"><div class=\"Lctrl\"><\/div><div class=\"Rctrl\"><\/div><\/div><\/div><noscript><a href=\"https:\/\/youtu.be\/vGxZIS4J0Mg\" rel=\"nofollow noopener\" target=\"_blank\"><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FvGxZIS4J0Mg%2F0.jpg\" alt=\"LTX Director 2 + SEED HUNTER workflow | The Dangers of Convenience\/Overhaul Nodes\" width=\"853\" height=\"460\" \/><br \/>Watch this video on YouTube<\/a><\/noscript><meta itemprop=\"description\" content=\"LTX DIRECTOR 2 SEED HUNTER WF: https:\/\/civitai.com\/models\/2720649\/ltx-director-2-seed-hunter-workflow (or for non-civitAI users: https:\/\/pastes.io\/rSkkMJV7) As per my video, I recommend you use traditional workflows such as my other SEED HUNTER workflows below: Single input: https:\/\/civitai.com\/models\/2676452\/ltx-23-seed-hunter-multiroll-workflow FFLF input: https:\/\/civitai.com\/models\/2688482\/ltx-23-fflf-seed-hunter-multiroll-workflow If you appreciate my work, please jump over to Patreon and follow for free (or not!): https:\/\/www.patreon.com\/c\/foxfuressence\"><\/div><\/div><div class=\"lL\" style=\"max-width:100%;width:853px;margin:5px auto;\"><\/div><figcaption><\/figcaption><\/figure>\n\n<figure class=\"wp-block-embed-youtube wp-block-embed is-type-video is-provider-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"lyte-wrapper\" title=\"\u591a\u5e27\u53c2\u8003+\u63d0\u793a\u8bcd\u63a5\u529b\uff1a\u63ed\u79d8 LTX Director \u5982\u4f55\u5b9e\u73b0\u597d\u83b1\u575e\u7ea7\u8fd0\u955c\u63a7\u5236|\u73a9\u8f6c LTX Director \u7684 5 \u4e2a\u9ad8\u9636\u6280\u5de7\uff1a\u8ba9\u4f60\u7684 AI \u89c6\u9891\u544a\u522b\u50f5\u786c\u4e0e\u95ea\u70c1\" style=\"width:853px;max-width:100%;margin:5px auto;\"><div class=\"lyMe\" id=\"WYL_IAr8rd9q4ZE\" itemprop=\"video\" itemscope itemtype=\"https:\/\/schema.org\/VideoObject\"><div><meta itemprop=\"thumbnailUrl\" content=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FIAr8rd9q4ZE%2Fhqdefault.jpg\" \/><meta itemprop=\"embedURL\" content=\"https:\/\/www.youtube.com\/embed\/IAr8rd9q4ZE\" \/><meta itemprop=\"duration\" content=\"PT12M32S\" \/><meta itemprop=\"uploadDate\" content=\"2026-05-25T05:49:02Z\" \/><\/div><meta itemprop=\"accessibilityFeature\" content=\"captions\" \/><div id=\"lyte_IAr8rd9q4ZE\" data-src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FIAr8rd9q4ZE%2Fhqdefault.jpg\" class=\"pL\"><div class=\"tC\"><div class=\"tT\" itemprop=\"name\">\u591a\u5e27\u53c2\u8003+\u63d0\u793a\u8bcd\u63a5\u529b\uff1a\u63ed\u79d8 LTX Director \u5982\u4f55\u5b9e\u73b0\u597d\u83b1\u575e\u7ea7\u8fd0\u955c\u63a7\u5236|\u73a9\u8f6c LTX Director \u7684 5 \u4e2a\u9ad8\u9636\u6280\u5de7\uff1a\u8ba9\u4f60\u7684 AI \u89c6\u9891\u544a\u522b\u50f5\u786c\u4e0e\u95ea\u70c1<\/div><\/div><div class=\"play\"><\/div><div class=\"ctrl\"><div class=\"Lctrl\"><\/div><div class=\"Rctrl\"><\/div><\/div><\/div><noscript><a href=\"https:\/\/youtu.be\/IAr8rd9q4ZE\" rel=\"nofollow noopener\" target=\"_blank\"><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FIAr8rd9q4ZE%2F0.jpg\" alt=\"\u591a\u5e27\u53c2\u8003+\u63d0\u793a\u8bcd\u63a5\u529b\uff1a\u63ed\u79d8 LTX Director \u5982\u4f55\u5b9e\u73b0\u597d\u83b1\u575e\u7ea7\u8fd0\u955c\u63a7\u5236|\u73a9\u8f6c LTX Director \u7684 5 \u4e2a\u9ad8\u9636\u6280\u5de7\uff1a\u8ba9\u4f60\u7684 AI \u89c6\u9891\u544a\u522b\u50f5\u786c\u4e0e\u95ea\u70c1\" width=\"853\" height=\"460\" \/><br \/>Watch this video on YouTube<\/a><\/noscript><meta itemprop=\"description\" content=\"\u4f60\u7684 AI \u89c6\u9891\u603b\u662f\u4e0d\u542c\u6307\u6325\u5417\uff1f\ud83c\udfac \u6b22\u8fce\u8ba4\u8bc6 LTX Director\uff0c\u8fd9\u662f LTX 2.3 \u4f53\u7cfb\u4e2d\u6700\u5f3a\u5927\u7684\u201c\u5168\u80fd\u578b\u201d\u8282\u70b9\uff0c\u5b83\u76f4\u63a5\u628a ComfyUI \u53d8\u6210\u4e86\u4e00\u4e2a\u4e13\u4e1a\u7684\u5bfc\u6f14\u63a7\u573a\u53f0\uff01 \u672c\u671f\u89c6\u9891\u6211\u5c06\u4e3a\u4f60\u6df1\u5ea6\u62c6\u89e3 Prompt Relay (\u63d0\u793a\u8bcd\u63a5\u529b) \u4e0e LTX Director \u7684\u6838\u5fc3\u8054\u7cfb\uff0c\u6559\u4f60\u5982\u4f55\u5c06\u4e0d\u540c\u7684\u6307\u4ee4\u7cbe\u51c6\u5206\u914d\u5230\u89c6\u9891\u7684\u4e0d\u540c\u7247\u6bb5\u3002\u6700\u91cd\u8981\u7684\u662f\uff0c\u5b98\u65b9\u63d0\u4f9b\u7684\u5de5\u4f5c\u6d41\u6548\u679c\u5f80\u5f80\u5dee\u5f3a\u4eba\u610f\u2014\u2014\u56e0\u6b64\u6211\u4eb2\u81ea\u6253\u9020\u4e86\u4e00\u5957**\u201cLTX Director \u6539\u8fdb\u7248\u5de5\u4f5c\u6d41\u201d**\uff0c\u901a\u8fc7 ManualSigmas \u91c7\u6837\u4f18\u5316\uff0c\u663e\u8457\u63d0\u5347\u4e86\u52a8\u4f5c\u5e45\u5ea6\u548c\u753b\u9762\u8d28\u611f\u3002 \ud83d\udd25 \u672c\u671f\u5e72\u8d27\u5185\u5bb9\u5168\u89e3\u6790\uff1a \u2022 \ud83e\udde0 \u5bfc\u6f14\u601d\u7ef4\u903b\u8f91\uff1a \u8be6\u89e3\u201c\u5168\u5c40\u63d0\u793a\u8bcd\u201d\u9501\u5b9a\u8eab\u4efd\u4e0e\u201c\u5c40\u90e8\u63d0\u793a\u8bcd\u201d\u5207\u6362\u52a8\u4f5c\u7684\u5e73\u8861\u672f\u3002 \u2022 \ud83d\udee0\ufe0f \u5de5\u4f5c\u6d41\u5927\u6539\u9020\uff1a \u63ed\u79d8\u5b98\u65b9\u8bbe\u7f6e\u7684\u5c40\u9650\uff0c\u6f14\u793a\u6539\u8fdb\u7248\u5982\u4f55\u8ba9\u52a8\u4f5c\u66f4\u6d41\u7545\u3001\u7ec6\u8282\u66f4\u624e\u5b9e\u3002 \u2022 \ud83c\udfad \u65e0\u5ba1\u67e5\u6a21\u578b\u8054\u52a8\uff1a \u5b9e\u6d4b\u914d\u5408 Sulphur \u6a21\u578b\uff0c\u8ba9\u4eba\u7269\u52a8\u4f5c\u548c\u955c\u5934\u4e92\u52a8\u66f4\u52a0\u5927\u80c6\u653e\u5f97\u5f00\u3002 \u2022 \ud83c\udf9e\ufe0f \u591a\u5e27\u63a7\u5236\u6280\u5de7\uff1a \u624b\u628a\u624b\u6559\u4f60\u9996\u5c3e\u5e27\u53c2\u8003\u3001\u591a\u4e2d\u95f4\u5e27\u5e03\u5c40\uff0c\u4ee5\u53ca\u5982\u4f55\u907f\u5f00\u591a\u5e27\u53c2\u8003\u5bfc\u81f4\u7684\u201c\u7ffb\u8f66\u201d\u9677\u9631\u3002 \u2022 \u23f1\ufe0f \u8282\u594f\u638c\u63a7\u529b\uff1a \u5982\u4f55\u6839\u636e\u52a8\u4f5c\u590d\u6742\u5ea6\u7cbe\u51c6\u5206\u914d\u5e27\u6570\uff0c\u5b9e\u73b0\u7535\u5f71\u7ea7\u7684\u526a\u8f91\u8282\u594f\u3002 #runninghub #LTXDirector #LTX23 #ComfyUI #AIVideo #PromptRelay #AIAnimation #VideoGeneration #AIFilmmaking #RunningHub #WhatDreamsCost #ManualSigmas #SulphurAI #AITutorial #GenerativeAI #VideoTimeline #AI\u7ed8\u753b #AI\u89c6\u9891 #\u5bfc\u6f14\u6a21\u5f0f #\u63d0\u793a\u8bcd\u63a5\u529b #\u5de5\u4f5c\u6d41\u5206\u4eab #\u4eba\u5de5\u667a\u80fd ------------------------------------------------------------------------------------------------------------------------ \u76f8\u5173\u8d44\u6e90 https:\/\/github.com\/WhatDreamsCost\/WhatDreamsCost-ComfyUI \u5728\u7ebf\u5de5\u4f5c\u6d41 1\uff1aLTX Director-i2v-prompt_replay-fixed \u4f53\u9a8c\u5730\u5740(\u56fd\u5185)\uff1ahttps:\/\/www.runninghub.cn\/post\/2057141350105178114\/?inviteCode=kol01-rh107 \u4f53\u9a8c\u5730\u5740(\u56fd\u9645)\uff1ahttps:\/\/www.runninghub.ai\/post\/2057141350105178114\/?inviteCode=kol02-rh021 \u5de5\u4f5c\u6d41\u4e0e\u63d0\u793a\u8bcd\u5de5\u7a0b\u4e0b\u8f7d https:\/\/github.com\/amao2001\/ganloss-latent-space\/blob\/main\/workflow\/2026-5-25LTXDirector-i2v-prompt_replay-fixed.json runninghub\u7c89\u4e1d\u798f\u5229 \u6ce8\u518c\u90011000\u70b9\u7b97\u529b \u56fd\u5916\u7528\u6237 https:\/\/www.runninghub.ai\/?utm_source=kol02-RH021 \u56fd\u5185\u7528\u6237 https:\/\/www.runninghub.cn\/?utm_source=kol01-RH107 ------------------------------------------------------------------------------------------------------------------------ \u80cc\u666f\u97f3\u4e50\u7531 AI \u751f\u6210\"><\/div><\/div><div class=\"lL\" style=\"max-width:100%;width:853px;margin:5px auto;\"><\/div><figcaption><\/figcaption><\/figure>\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<h2 class=\"wp-block-heading\">LTX IC-LoRA<\/h2>\n\n\n<figure class=\"wp-block-embed-youtube wp-block-embed is-type-video is-provider-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"lyte-wrapper\" title=\"I Tested the New LTX 2.3 Ingredients LoRA &mdash; This Is Huge for AI Filmmaking\" style=\"width:853px;max-width:100%;margin:5px auto;\"><div class=\"lyMe\" id=\"WYL_guxp6uFqwn8\" itemprop=\"video\" itemscope itemtype=\"https:\/\/schema.org\/VideoObject\"><div><meta itemprop=\"thumbnailUrl\" content=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2Fguxp6uFqwn8%2Fhqdefault.jpg\" \/><meta itemprop=\"embedURL\" content=\"https:\/\/www.youtube.com\/embed\/guxp6uFqwn8\" \/><meta itemprop=\"duration\" content=\"PT12M21S\" \/><meta itemprop=\"uploadDate\" content=\"2026-07-02T13:02:30Z\" \/><\/div><div id=\"lyte_guxp6uFqwn8\" data-src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2Fguxp6uFqwn8%2Fhqdefault.jpg\" class=\"pL\"><div class=\"tC\"><div class=\"tT\" itemprop=\"name\">I Tested the New LTX 2.3 Ingredients LoRA \u2014 This Is Huge for AI Filmmaking<\/div><\/div><div class=\"play\"><\/div><div class=\"ctrl\"><div class=\"Lctrl\"><\/div><div class=\"Rctrl\"><\/div><\/div><\/div><noscript><a href=\"https:\/\/youtu.be\/guxp6uFqwn8\" rel=\"nofollow noopener\" target=\"_blank\"><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2Fguxp6uFqwn8%2F0.jpg\" alt=\"I Tested the New LTX 2.3 Ingredients LoRA &mdash; This Is Huge for AI Filmmaking\" width=\"853\" height=\"460\" \/><br \/>Watch this video on YouTube<\/a><\/noscript><meta itemprop=\"description\" content=\"In this video, I show you the new LTX 2.3 Ingredients IC-LoRA \u2014 one of the most interesting tools for creating more consistent AI video scenes inside ComfyUI. This LoRA is not just another style LoRA. It works with a reference sheet, where you can prepare all the important \u201cingredients\u201d of your scene: characters, outfits, props, objects, locations, colors, lighting, and visual style. Then LTX 2.3 uses those references to generate a new cinematic video while trying to preserve consistency across the entire shot. This is a big step forward for AI filmmaking, because one of the hardest problems in AI video is keeping the same characters, faces, clothes, props, and environments stable from frame to frame. In this tutorial, I show: How the LTX 2.3 Ingredients LoRA works How to prepare a proper reference sheet How to use characters, props, outfits, and locations as scene ingredients How to connect the workflow in ComfyUI How to generate a cinematic AI video scene Why this can be useful for animated series, short films, product videos, and local AI video production You can download the workflow, model links, and reference images from my Patreon post here: https:\/\/www.patreon.com\/AntiMatterComfy\/posts\/ltx-2-3-lora-in-162691572?utm_medium=clipboard_copy&amp;utm_source=copyLink&amp;utm_campaign=postshare_creator&amp;utm_content=join_link This is a free workflow post, so you can test it locally and experiment with your own characters and scenes. If you want more advanced system prompts, agent instructions, and professional prompt setups for generating better Ingredients LoRA reference sheets and video prompts, check out my paid Patreon tier. Thanks for watching, and subscribe for more ComfyUI, LTX 2.3, AI video, and local AI filmmaking workflows\"><\/div><\/div><div class=\"lL\" style=\"max-width:100%;width:853px;margin:5px auto;\"><\/div><figcaption><\/figcaption><\/figure>\n\n<figure class=\"wp-block-embed-youtube wp-block-embed is-type-video is-provider-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"lyte-wrapper\" title=\"LTX 2.3 Ingredients IC-LoRA &mdash; The NEW Reference-to-Video Model\" style=\"width:853px;max-width:100%;margin:5px auto;\"><div class=\"lyMe\" id=\"WYL_P-nRo5muRCM\" itemprop=\"video\" itemscope itemtype=\"https:\/\/schema.org\/VideoObject\"><div><meta itemprop=\"thumbnailUrl\" content=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FP-nRo5muRCM%2Fhqdefault.jpg\" \/><meta itemprop=\"embedURL\" content=\"https:\/\/www.youtube.com\/embed\/P-nRo5muRCM\" \/><meta itemprop=\"duration\" content=\"PT9M42S\" \/><meta itemprop=\"uploadDate\" content=\"2026-06-18T13:57:26Z\" \/><\/div><div id=\"lyte_P-nRo5muRCM\" data-src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FP-nRo5muRCM%2Fhqdefault.jpg\" class=\"pL\"><div class=\"tC\"><div class=\"tT\" itemprop=\"name\">LTX 2.3 Ingredients IC-LoRA \u2014 The NEW Reference-to-Video Model<\/div><\/div><div class=\"play\"><\/div><div class=\"ctrl\"><div class=\"Lctrl\"><\/div><div class=\"Rctrl\"><\/div><\/div><\/div><noscript><a href=\"https:\/\/youtu.be\/P-nRo5muRCM\" rel=\"nofollow noopener\" target=\"_blank\"><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FP-nRo5muRCM%2F0.jpg\" alt=\"LTX 2.3 Ingredients IC-LoRA &mdash; The NEW Reference-to-Video Model\" width=\"853\" height=\"460\" \/><br \/>Watch this video on YouTube<\/a><\/noscript><meta itemprop=\"description\" content=\"Lightricks just dropped a major update to the LTX 2.3 ecosystem: a completely rebuilt LoRA trainer with native IC-LoRA support, and a brand-new &quot;Ingredients&quot; model for reference-to-video generation. In this video, we walk through the full new trainer pipeline, show you how to use the Ingredients IC-LoRA to generate multi-character video from a single reference sheet, and compare it head-to-head with Lykon Studio&#039;s MSR-LoRA. We also cover the practical ComfyUI workflow, why Prompt Relay doesn&#039;t work with Ingredients right now, and how to set up your own reference sheet using Ideogram 4 and AI-assisted prompt generation. This content is built for AI video creators, ComfyUI users, and anyone working with the LTX model family. Whether you&#039;re generating multi-character scenes, training your own video LoRAs, or just trying to figure out which reference-to-video approach is best for your workflow, this video gives you a hands-on breakdown with real examples. No fluff \u2014 just what works, what doesn&#039;t, and how to get the best results on consumer hardware. This matters because the new LTX trainer opens the door for anyone to train custom video LoRAs, not just rely on community releases. With built-in agent instructions for AI-assisted training, support for audio and video conditioning, and a growing Creative Lab collection, LTX 2.3 is becoming a serious production pipeline. Understanding the Ingredients model now \u2014 and its strengths and limitations compared to alternatives like MSR-LoRA \u2014 will help you pick the right tool for your projects and prepare for the next wave of custom-trained models. News About New LTX Trainer https:\/\/x.com\/ltx_io\/status\/2067284780081905685 Github https:\/\/github.com\/Lightricks\/LTX-2 Model https:\/\/huggingface.co\/Lightricks\/LTX-2.3 LoRAs and IC-LoRAs, trained on the LTX-2.3 model https:\/\/huggingface.co\/collections\/Lightricks\/ltx-23-creative-lab Lightricks\/LTX-2.3-22b-IC-LoRA-Ingredients https:\/\/huggingface.co\/Lightricks\/LTX-2.3-22b-IC-LoRA-Ingredients Mod Workflow based on Lightricks Github workflow https:\/\/www.patreon.com\/aifuturetech\/posts\/ltx-2-3-ic-lora-161431864 Local Workstation GPU : https:\/\/amzn.to\/3XfXsAO -------------------------------------------------------------------------------------------------------------------------------- If You Like tutorial like this, You Can Support Our Work In Patreon: https:\/\/www.patreon.com\/c\/aifuturetech\"><\/div><\/div><div class=\"lL\" style=\"max-width:100%;width:853px;margin:5px auto;\"><\/div><figcaption><\/figcaption><\/figure>\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<h2 class=\"wp-block-heading\">LTX 2.3<\/h2>\n\n\n<figure class=\"wp-block-embed-youtube wp-block-embed is-type-video is-provider-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"lyte-wrapper\" title=\"I Made Elon Musk Do This... (Free Open-Source AI Video)\" style=\"width:853px;max-width:100%;margin:5px auto;\"><div class=\"lyMe\" id=\"WYL_AtEstFHy3Dc\" itemprop=\"video\" itemscope itemtype=\"https:\/\/schema.org\/VideoObject\"><div><meta itemprop=\"thumbnailUrl\" content=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FAtEstFHy3Dc%2Fhqdefault.jpg\" \/><meta itemprop=\"embedURL\" content=\"https:\/\/www.youtube.com\/embed\/AtEstFHy3Dc\" \/><meta itemprop=\"duration\" content=\"PT10M9S\" \/><meta itemprop=\"uploadDate\" content=\"2026-06-20T02:25:31Z\" \/><\/div><div id=\"lyte_AtEstFHy3Dc\" data-src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FAtEstFHy3Dc%2Fhqdefault.jpg\" class=\"pL\"><div class=\"tC\"><div class=\"tT\" itemprop=\"name\">I Made Elon Musk Do This... (Free Open-Source AI Video)<\/div><\/div><div class=\"play\"><\/div><div class=\"ctrl\"><div class=\"Lctrl\"><\/div><div class=\"Rctrl\"><\/div><\/div><\/div><noscript><a href=\"https:\/\/youtu.be\/AtEstFHy3Dc\" rel=\"nofollow noopener\" target=\"_blank\"><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FAtEstFHy3Dc%2F0.jpg\" alt=\"I Made Elon Musk Do This... (Free Open-Source AI Video)\" width=\"853\" height=\"460\" \/><br \/>Watch this video on YouTube<\/a><\/noscript><meta itemprop=\"description\" content=\"Can open-source AI video finally handle flawless character consistency? Today, we are stress-testing the brand new LTX 2.3 Multiple Subject Reference (MSR) model by LiconStudio. I put this 22-billion parameter architecture to the absolute test using multiple characters, custom props, and chaotic prompt changes to see if it actually lives up to the developer hype or falls apart under pressure. Best of all, I\u2019ve put together a 100% free Kaggle notebook so you can run this model yourself without needing a high-end local GPU. We\u2019ll break down the token-level visual memory system, look at the best settings to prevent your generations from breaking, and look at the brutal truth regarding its identity retention limitations. \u25bd\u00a0TOOLS &amp; RESOURCES MENTIONED\u00a0\u25bd \u25ba My Free Notebooks Collections: https:\/\/github.com\/TeamAIQ\/Colab-notebooks \u25ba LTX 2.3 MSR Hugging Face Page: https:\/\/huggingface.co\/LiconStudio\/LTX-2.3-Multiple-Subject-Reference \u25ac\u25ac\u25ac\u25ac\u25ac\u25ac\u25ac\u25ac\u25ac\u25ac\u25ac\u25ac\u25ac\u25ac\u25ac\u25ac\u25ac\u25ac\u25ac\u25ac\u25ac \ud83e\udd1d Business: TeamAIQ@yahoo.com \ud83d\ude80 X \/ Twitter: https:\/\/x.com\/AiquestAcademy \u2615 My bio link: https:\/\/aiquest.site\/ \u25ac\u25ac\u25ac\u25ac\u25ac\u25ac\u25ac\u25ac\u25ac\u25ac\u25ac\u25ac\u25ac\u25ac\u25ac\u25ac\u25ac\u25ac\u25ac\u25ac\u25ac #AIVideo #OpenSourceAI #CharacterConsistency\"><\/div><\/div><div class=\"lL\" style=\"max-width:100%;width:853px;margin:5px auto;\"><\/div><figcaption><\/figcaption><\/figure>\n\n<figure class=\"wp-block-embed-youtube wp-block-embed is-type-video is-provider-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"lyte-wrapper\" title=\"LTX2.3\u4e00\u952e\u6574\u5408\u5305\uff0c\u6570\u5b57\u4eba\u5bf9\u53e3\u578b\uff0cAI\u5531\u6b4c\uff0c\u9996\u5c3e\u5e27\uff0c\u56fe\u751f\u89c6\u9891\uff0c\u6587\u751f\u89c6\u9891\uff0c\u56fe\u7247\u7f16\u8f91\uff0c\u6587\u751f\u56fe\uff0c\u652f\u630150\u7cfb\u663e\u5361\uff0c8G\u663e\u5b58\u53ef\u7528\uff0c64G\u5185\u5b58\u3002\" style=\"width:853px;max-width:100%;margin:5px auto;\"><div class=\"lyMe\" id=\"WYL_aGHew8uDbZs\" itemprop=\"video\" itemscope itemtype=\"https:\/\/schema.org\/VideoObject\"><div><meta itemprop=\"thumbnailUrl\" content=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FaGHew8uDbZs%2Fhqdefault.jpg\" \/><meta itemprop=\"embedURL\" content=\"https:\/\/www.youtube.com\/embed\/aGHew8uDbZs\" \/><meta itemprop=\"duration\" content=\"PT6M9S\" \/><meta itemprop=\"uploadDate\" content=\"2026-06-13T15:32:59Z\" \/><\/div><meta itemprop=\"accessibilityFeature\" content=\"captions\" \/><div id=\"lyte_aGHew8uDbZs\" data-src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FaGHew8uDbZs%2Fhqdefault.jpg\" class=\"pL\"><div class=\"tC\"><div class=\"tT\" itemprop=\"name\">LTX2.3\u4e00\u952e\u6574\u5408\u5305\uff0c\u6570\u5b57\u4eba\u5bf9\u53e3\u578b\uff0cAI\u5531\u6b4c\uff0c\u9996\u5c3e\u5e27\uff0c\u56fe\u751f\u89c6\u9891\uff0c\u6587\u751f\u89c6\u9891\uff0c\u56fe\u7247\u7f16\u8f91\uff0c\u6587\u751f\u56fe\uff0c\u652f\u630150\u7cfb\u663e\u5361\uff0c8G\u663e\u5b58\u53ef\u7528\uff0c64G\u5185\u5b58\u3002<\/div><\/div><div class=\"play\"><\/div><div class=\"ctrl\"><div class=\"Lctrl\"><\/div><div class=\"Rctrl\"><\/div><\/div><\/div><noscript><a href=\"https:\/\/youtu.be\/aGHew8uDbZs\" rel=\"nofollow noopener\" target=\"_blank\"><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FaGHew8uDbZs%2F0.jpg\" alt=\"LTX2.3\u4e00\u952e\u6574\u5408\u5305\uff0c\u6570\u5b57\u4eba\u5bf9\u53e3\u578b\uff0cAI\u5531\u6b4c\uff0c\u9996\u5c3e\u5e27\uff0c\u56fe\u751f\u89c6\u9891\uff0c\u6587\u751f\u89c6\u9891\uff0c\u56fe\u7247\u7f16\u8f91\uff0c\u6587\u751f\u56fe\uff0c\u652f\u630150\u7cfb\u663e\u5361\uff0c8G\u663e\u5b58\u53ef\u7528\uff0c64G\u5185\u5b58\u3002\" width=\"853\" height=\"460\" \/><br \/>Watch this video on YouTube<\/a><\/noscript><meta itemprop=\"description\" content=\"1\u3001\u6548\u679c\u975e\u5e38\u60ca\u8273 LTX2.3\uff1a\u6570\u5b57\u4eba\uff0c\u56fe\u7247\u3001\u6587\u5b57\u751f\u6210\u89c6\u9891\uff0c\u6574\u5408\u5305 + \u4e91\u7aef\u6211\u90fd\u505a\u597d\u4e86\uff01 2\u3001\u4e91\u7aef\u4f53\u9a8c\uff1ahttps:\/\/www.compshare.cn\/images\/gbEUSxYJatWb?referral_code=bZRguxQSRDBJHgp5vAE7u&amp;ytag=GPU_YY_YX_bl_yuziyuetalk0614 3\u3001\u6574\u5408\u5305\uff1a\u5173\u6ce8 + \u4e09\u8fde \u53d1\u9001\u79c1\u4fe1\u5173\u952e\u5b57 \u201cltx23\u201d \u81ea\u53d6 \u65f6\u95f4\u8f74\uff1a 00:00 \u5f00\u573a 01:44 \u6574\u5408\u5305 04:49 \u663e\u5b58\u5360\u7528 05:09 \u4e91\u7aef\u4f7f\u7528\"><\/div><\/div><div class=\"lL\" style=\"max-width:100%;width:853px;margin:5px auto;\"><\/div><figcaption><\/figcaption><\/figure>\n\n<figure class=\"wp-block-embed-youtube wp-block-embed is-type-video is-provider-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"lyte-wrapper\" title=\"LTX 2.3 Character Motion Transfer | Video to Image Animation\" style=\"width:853px;max-width:100%;margin:5px auto;\"><div class=\"lyMe\" id=\"WYL_JiZ6GX_HwPg\" itemprop=\"video\" itemscope itemtype=\"https:\/\/schema.org\/VideoObject\"><div><meta itemprop=\"thumbnailUrl\" content=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FJiZ6GX_HwPg%2Fhqdefault.jpg\" \/><meta itemprop=\"embedURL\" content=\"https:\/\/www.youtube.com\/embed\/JiZ6GX_HwPg\" \/><meta itemprop=\"duration\" content=\"PT10M30S\" \/><meta itemprop=\"uploadDate\" content=\"2026-06-12T12:14:16Z\" \/><\/div><meta itemprop=\"accessibilityFeature\" content=\"captions\" \/><div id=\"lyte_JiZ6GX_HwPg\" data-src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FJiZ6GX_HwPg%2Fhqdefault.jpg\" class=\"pL\"><div class=\"tC\"><div class=\"tT\" itemprop=\"name\">LTX 2.3 Character Motion Transfer | Video to Image Animation<\/div><\/div><div class=\"play\"><\/div><div class=\"ctrl\"><div class=\"Lctrl\"><\/div><div class=\"Rctrl\"><\/div><\/div><\/div><noscript><a href=\"https:\/\/youtu.be\/JiZ6GX_HwPg\" rel=\"nofollow noopener\" target=\"_blank\"><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FJiZ6GX_HwPg%2F0.jpg\" alt=\"LTX 2.3 Character Motion Transfer | Video to Image Animation\" width=\"853\" height=\"460\" \/><br \/>Watch this video on YouTube<\/a><\/noscript><meta itemprop=\"description\" content=\"\ud83c\udf1fLTX Prompt Timeline: Turn 5s Clips into Full Videos\ud83c\udf1f In this video, I demonstrate a character motion transfer workflow using LTX 2.3 and a LoRA. The workflow takes: A reference image of a character A driving video and transfers the motion from the video to the character in the image, creating an animated result. I also explain the workflow and share the files used so you can try it yourself. #LTX23 #LTXVideo #AIAnimation #MotionTransfer #ComfyUI #LoRA #AIVideo #CharacterAnimation #OpenSourceAI #GenerativeAI \ud83d\udc49 I\u2019ve shared two examples to demonstrate how scene changes, camera motion, and continuity can be controlled using this approach. \ud83d\ude80 Resources: Workflow, image and video I have used. https:\/\/discord.com\/channels\/1414348074477949022\/1514957453249347716 \u2728 https:\/\/huggingface.co\/datasets\/Yogesh-DevHub\/LTX2.3\/tree\/main\/LTX-IV2V-Transfer-Motion Install Sage Attention Video - 1: https:\/\/youtu.be\/-S39owjSsMo Video - 2: https:\/\/youtu.be\/b43GLxkbg6o Stay tuned for the next update \ud83d\udc40 #comfyui #aivideo #aiimage #aitools \ud83d\udcbb Hardware Used: (i5 14400F, 64GB RAM, RTX 5060 TI 16GB) \ud83d\udce5 Get Started with Comfy UI: https:\/\/youtu.be\/grzK5mBitzs \ud83d\udcac Drop a comment below! \ud83d\udd14 Stay Updated: Stay connected to find the latest AI applications! Happy creating! \ud83c\udfa8\u2728 Playlist For ComfyUI - Stable Diffusion https:\/\/www.youtube.com\/playlist?list=PLPFN04WspxqvpVeMmJ1vHqG-uf4C4dIb0 Playlist For Krita AI Diffusion https:\/\/www.youtube.com\/playlist?list=PLPFN04WspxqvFhJDIXvIDZ3yveShMvgss Playlist For Fooocus AI - Stable Diffusion https:\/\/www.youtube.com\/playlist?list=PLPFN04WspxqsslRSpiLmwGR8QTpDYNv7z Playlist For Forge UI https:\/\/www.youtube.com\/playlist?list=PLPFN04WspxqvpZmxPPjkxV-BkuOy2FF2i\"><\/div><\/div><div class=\"lL\" style=\"max-width:100%;width:853px;margin:5px auto;\"><\/div><figcaption><\/figcaption><\/figure>\n\n<figure class=\"wp-block-embed-youtube wp-block-embed is-type-video is-provider-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"lyte-wrapper\" title=\"BERNINI ! Video + Image Ref 2 Video editor in ComyUI (DEEP DIVE)\" style=\"width:853px;max-width:100%;margin:5px auto;\"><div class=\"lyMe\" id=\"WYL_5oQ4F4G9MaE\" itemprop=\"video\" itemscope itemtype=\"https:\/\/schema.org\/VideoObject\"><div><meta itemprop=\"thumbnailUrl\" content=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2F5oQ4F4G9MaE%2Fhqdefault.jpg\" \/><meta itemprop=\"embedURL\" content=\"https:\/\/www.youtube.com\/embed\/5oQ4F4G9MaE\" \/><meta itemprop=\"duration\" content=\"PT17M14S\" \/><meta itemprop=\"uploadDate\" content=\"2026-06-07T15:56:54Z\" \/><\/div><div id=\"lyte_5oQ4F4G9MaE\" data-src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2F5oQ4F4G9MaE%2Fhqdefault.jpg\" class=\"pL\"><div class=\"tC\"><div class=\"tT\" itemprop=\"name\">BERNINI ! Video + Image Ref 2 Video editor in ComyUI (DEEP DIVE)<\/div><\/div><div class=\"play\"><\/div><div class=\"ctrl\"><div class=\"Lctrl\"><\/div><div class=\"Rctrl\"><\/div><\/div><\/div><noscript><a href=\"https:\/\/youtu.be\/5oQ4F4G9MaE\" rel=\"nofollow noopener\" target=\"_blank\"><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2F5oQ4F4G9MaE%2F0.jpg\" alt=\"BERNINI ! Video + Image Ref 2 Video editor in ComyUI (DEEP DIVE)\" width=\"853\" height=\"460\" \/><br \/>Watch this video on YouTube<\/a><\/noscript><meta itemprop=\"description\" content=\"BRILLIANT ! Bernini can edit an existing video by using an image reference to change the video. Here we look at Bernini step by step with real results If you liked this video, please consider clicking the LIKE button and subscribing, thanks. LEAVE A COMMENT. Join our &quot;Get Going Fast&quot; Discord server here for the workflow and chat https:\/\/discord.gg\/nn3RRcVy5c For one-click installs to Get Going Fast visit The AI Hobby Guy here and create a FREE account https:\/\/getgoingfast.pro\/ #bernini #wan22 #zimage #firered #comfyui #bernini\"><\/div><\/div><div class=\"lL\" style=\"max-width:100%;width:853px;margin:5px auto;\"><\/div><figcaption><\/figcaption><\/figure>\n\n<figure class=\"wp-block-embed-youtube wp-block-embed is-type-video is-provider-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"lyte-wrapper\" title=\"How to Use Exciting LTX 2.3 LoRAs for Stunning AI Videos\" style=\"width:853px;max-width:100%;margin:5px auto;\"><div class=\"lyMe\" id=\"WYL_zA55q3zG930\" itemprop=\"video\" itemscope itemtype=\"https:\/\/schema.org\/VideoObject\"><div><meta itemprop=\"thumbnailUrl\" content=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FzA55q3zG930%2Fhqdefault.jpg\" \/><meta itemprop=\"embedURL\" content=\"https:\/\/www.youtube.com\/embed\/zA55q3zG930\" \/><meta itemprop=\"duration\" content=\"PT5M16S\" \/><meta itemprop=\"uploadDate\" content=\"2026-06-12T16:40:57Z\" \/><\/div><meta itemprop=\"accessibilityFeature\" content=\"captions\" \/><div id=\"lyte_zA55q3zG930\" data-src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FzA55q3zG930%2Fhqdefault.jpg\" class=\"pL\"><div class=\"tC\"><div class=\"tT\" itemprop=\"name\">How to Use Exciting LTX 2.3 LoRAs for Stunning AI Videos<\/div><\/div><div class=\"play\"><\/div><div class=\"ctrl\"><div class=\"Lctrl\"><\/div><div class=\"Rctrl\"><\/div><\/div><\/div><noscript><a href=\"https:\/\/youtu.be\/zA55q3zG930\" rel=\"nofollow noopener\" target=\"_blank\"><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FzA55q3zG930%2F0.jpg\" alt=\"How to Use Exciting LTX 2.3 LoRAs for Stunning AI Videos\" width=\"853\" height=\"460\" \/><br \/>Watch this video on YouTube<\/a><\/noscript><meta itemprop=\"description\" content=\"\ud83d\udd25 In this video, we&#039;re exploring some of the most interesting LoRA models available for LTX 2.3 and how they can transform ordinary generations into something truly unique. I&#039;ll also show you how I use an LTX Director workflow with a low-VRAM GGUF model and OmniNFT RL-LoRA to achieve better performance, improved lip-sync, and more natural character animations. \u2728 We&#039;ll dive into Cinematic Post Apocalyptic, a LoRA that turns simple footage into gritty, photorealistic survival scenes straight out of a blockbuster movie. \u2728 Next, we&#039;ll take a look at LTX2.3-IC-LORA-Dual-Character, a powerful solution for maintaining character consistency in conversations and action scenes\u2014one of the biggest challenges in AI video generation. \u2728 Need a break from filmmaking? We&#039;ll switch gears and experiment with K-pop Dance LoRA, complete with its quirks, unexpected results, and a few lessons learned through trial and error. \u2728 We&#039;ll also explore Latin Dance LoRA, which pairs beautifully with Image-to-Video workflows and custom music tracks, allowing you to create energetic and personalized dance performances. \u2728 Finally, we&#039;ll step into the neon-lit world of Synthwave LoRA, where retro-futurism, 1980s nostalgia, and vibrant aesthetics come together to create visually stunning videos. Whether you&#039;re making cinematic stories, dance videos, or experimenting with new visual styles, I hope this video gives you fresh ideas and inspiration for your next AI project. If you found this video helpful, please consider leaving a like and subscribing to the channel. Your support means a lot and helps me continue creating more AI tutorials and creative experiments. Thanks for watching. Keep creating, keep exploring, and I&#039;ll see you in the next AI adventure! \u2728 All workflows and resources are shared for free. If you\u2019d like to support the channel, you can buy me a coffee \u2764\ufe0f: https:\/\/ko-fi.com\/fututek -------------------------------------------------- \u2728Run on Cloud GPU: https:\/\/runpod.io?ref=99cuxxw1 -------------------------------------------------- \u2728Workflow and resources: https:\/\/drive.google.com\/drive\/folders\/1yxMVGqgDcJDIqSE_X61YQOdQxHLEHMGx?usp=sharing - LTX2.3 Model for low VRAM 8~12Gb (Place in \/ComfyUI\/models\/diffusion_models) https:\/\/huggingface.co\/vantagewithai\/LTX-2.3-GGUF\/blob\/main\/dev\/ltx-2-3-22b-dev-Q4_K_M.gguf - Text Encoder (Place in \/ComfyUI\/models\/text_encoders) https:\/\/huggingface.co\/Comfy-Org\/ltx-2\/resolve\/main\/split_files\/text_encoders\/gemma_3_12B_it_fp4_mixed.safetensors https:\/\/huggingface.co\/Kijai\/LTX2.3_comfy\/blob\/main\/text_encoders\/ltx-2.3_text_projection_bf16.safetensors - VAE (Place in \/ComfyUI\/models\/vae) https:\/\/huggingface.co\/Kijai\/LTX2.3_comfy\/blob\/main\/vae\/LTX23_audio_vae_bf16.safetensors https:\/\/huggingface.co\/Kijai\/LTX2.3_comfy\/blob\/main\/vae\/LTX23_video_vae_bf16.safetensors - Spatial Upscaler (Place in \/ComfyUI\/models\/latent_upscale_models) https:\/\/huggingface.co\/Lightricks\/LTX-2.3\/resolve\/main\/ltx-2.3-spatial-upscaler-x2-1.1.safetensors - Distilled Loras (Place in \/ComfyUI\/models\/loras) https:\/\/huggingface.co\/Kijai\/LTX2.3_comfy\/blob\/main\/loras\/ltx-2.3-22b-distilled-lora-dynamic_fro09_avg_rank_105_bf16.safetensors - Enhancing Loras (Place in \/ComfyUI\/models\/loras) 1. Cinematic Post-Apocalyptic: https:\/\/civitai.com\/models\/2563394\/ltx-23-cinematic-post-apocalyptic 2. Dual Characters: https:\/\/civitai.com\/models\/2500098\/ltx23-ic-lora-dual-character?modelVersionId=2810376 3. K-Pop Dance: https:\/\/civitai.com\/models\/2487939\/ltx-23-dance 4. Latin Rhythm: https:\/\/civitai.com\/models\/2649352\/ltx-23-latin-dance 5. Synthware https:\/\/civitai.com\/models\/2551439\/ltx-23-synthwave 6. OmniNFT RL-LoRA https:\/\/huggingface.co\/Kijai\/LTX2.3_comfy\/resolve\/main\/loras\/LTX-2.3-OmniNFT-RL-Lora_bf16.safetensors\"><\/div><\/div><div class=\"lL\" style=\"max-width:100%;width:853px;margin:5px auto;\"><\/div><figcaption><\/figcaption><\/figure>\n\n<figure class=\"wp-block-embed-youtube wp-block-embed is-type-video is-provider-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"lyte-wrapper\" title=\"This LTX-2.3 LoRA Removes Obstructions From Video | Wan2GP Tutorial\" style=\"width:853px;max-width:100%;margin:5px auto;\"><div class=\"lyMe\" id=\"WYL_1fOPyztY768\" itemprop=\"video\" itemscope itemtype=\"https:\/\/schema.org\/VideoObject\"><div><meta itemprop=\"thumbnailUrl\" content=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2F1fOPyztY768%2Fhqdefault.jpg\" \/><meta itemprop=\"embedURL\" content=\"https:\/\/www.youtube.com\/embed\/1fOPyztY768\" \/><meta itemprop=\"duration\" content=\"PT4M16S\" \/><meta itemprop=\"uploadDate\" content=\"2026-05-30T03:17:41Z\" \/><\/div><div id=\"lyte_1fOPyztY768\" data-src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2F1fOPyztY768%2Fhqdefault.jpg\" class=\"pL\"><div class=\"tC\"><div class=\"tT\" itemprop=\"name\">This LTX-2.3 LoRA Removes Obstructions From Video | Wan2GP Tutorial<\/div><\/div><div class=\"play\"><\/div><div class=\"ctrl\"><div class=\"Lctrl\"><\/div><div class=\"Rctrl\"><\/div><\/div><\/div><noscript><a href=\"https:\/\/youtu.be\/1fOPyztY768\" rel=\"nofollow noopener\" target=\"_blank\"><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2F1fOPyztY768%2F0.jpg\" alt=\"This LTX-2.3 LoRA Removes Obstructions From Video | Wan2GP Tutorial\" width=\"853\" height=\"460\" \/><br \/>Watch this video on YouTube<\/a><\/noscript><meta itemprop=\"description\" content=\"In this quick demo and tutorial, I show how to use Obscura Remova, an LTX-2.3 LoRA designed to remove obstructions, blocking objects, and unwanted foreground elements from video. We\u2019ll run the LoRA inside Wan2GP, test it on obstructed video examples, and compare the before\/after results to see how well LTX-2.3 can reconstruct a cleaner video frame. This is useful for AI video workflows where you want to remove fences, foreground objects, text, visual clutter, or other elements blocking the subject. \ud83d\udcda Resources: \ud83d\udd17 Setting Up Your Wan2GP Environment on RunPod: https:\/\/youtu.be\/_fZqf4bzMng \ud83d\udd17Obscura Remova LoRA: https:\/\/huggingface.co\/WepeNerd\/Obscura_Remova \ud83d\udd17 Wan2GP repo: https:\/\/github.com\/deepbeepmeep\/Wan2GP \ud83d\udcdd Topics covered: LTX-2.3 LoRA workflow Obscura Remova demo Wan2GP tutorial AI video object removal Removing obstructions from video Foreground object removal with AI video LTX 2.3 is an open video generation model used for image-to-video, video-to-video and text-to-video. Wan2GP is an open-source app for running AI image and video models on consumer GPUs, including LTX Video. It&#039;s like an easier to use ComfyUI. #LTX23 #LTXVideo #ObjectRemoval #Wan2GP #AIVideo #VideoEditing #AITools #opensourceai #runpod #AIFilmmaking #AIVideoGeneration #LoRA\"><\/div><\/div><div class=\"lL\" style=\"max-width:100%;width:853px;margin:5px auto;\"><\/div><figcaption><\/figcaption><\/figure>\n\n<figure class=\"wp-block-embed-youtube wp-block-embed is-type-video is-provider-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"lyte-wrapper\" title=\"\u3010ComfyUI\u3011\u5c0f\u5c3a\u5ea6\u5de5\u4f5c\u6d41\uff01\u628aLTX 2 3\u53c8\u69a8\u5e72\u4e86\u4e00\u6b21\uff1a\u89c6\u9891\u751f\u6210\u7ec8\u4e8e\u66f4\u7a33\u3001\u66f4\u51c6\u3001\u66f4\u542c\u8bdd\u4e86 p01 Ltx2 3\u7ee7\u7eed&ldquo;\u69a8\u5e72&rdquo;\" style=\"width:853px;max-width:100%;margin:5px auto;\"><div class=\"lyMe\" id=\"WYL_uQpRKwOjjZM\" itemprop=\"video\" itemscope itemtype=\"https:\/\/schema.org\/VideoObject\"><div><meta itemprop=\"thumbnailUrl\" content=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FuQpRKwOjjZM%2Fhqdefault.jpg\" \/><meta itemprop=\"embedURL\" content=\"https:\/\/www.youtube.com\/embed\/uQpRKwOjjZM\" \/><meta itemprop=\"duration\" content=\"PT4M19S\" \/><meta itemprop=\"uploadDate\" content=\"2026-05-22T02:36:43Z\" \/><\/div><div id=\"lyte_uQpRKwOjjZM\" data-src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FuQpRKwOjjZM%2Fhqdefault.jpg\" class=\"pL\"><div class=\"tC\"><div class=\"tT\" itemprop=\"name\">\u3010ComfyUI\u3011\u5c0f\u5c3a\u5ea6\u5de5\u4f5c\u6d41\uff01\u628aLTX 2 3\u53c8\u69a8\u5e72\u4e86\u4e00\u6b21\uff1a\u89c6\u9891\u751f\u6210\u7ec8\u4e8e\u66f4\u7a33\u3001\u66f4\u51c6\u3001\u66f4\u542c\u8bdd\u4e86 p01 Ltx2 3\u7ee7\u7eed\u201c\u69a8\u5e72\u201d<\/div><\/div><div class=\"play\"><\/div><div class=\"ctrl\"><div class=\"Lctrl\"><\/div><div class=\"Rctrl\"><\/div><\/div><\/div><noscript><a href=\"https:\/\/youtu.be\/uQpRKwOjjZM\" rel=\"nofollow noopener\" target=\"_blank\"><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FuQpRKwOjjZM%2F0.jpg\" alt=\"\u3010ComfyUI\u3011\u5c0f\u5c3a\u5ea6\u5de5\u4f5c\u6d41\uff01\u628aLTX 2 3\u53c8\u69a8\u5e72\u4e86\u4e00\u6b21\uff1a\u89c6\u9891\u751f\u6210\u7ec8\u4e8e\u66f4\u7a33\u3001\u66f4\u51c6\u3001\u66f4\u542c\u8bdd\u4e86 p01 Ltx2 3\u7ee7\u7eed&ldquo;\u69a8\u5e72&rdquo;\" width=\"853\" height=\"460\" \/><br \/>Watch this video on YouTube<\/a><\/noscript><meta itemprop=\"description\" content=\"\u300cLTX\u63d2\u4ef6\u548c\u6a21\u578b\u5168\u5957\u300d\u94fe\u63a5\uff1ahttps:\/\/pan.quark.cn\/s\/58b98f6d688f \u300cComfyUI V8\u5b8c\u6574\u6574\u5408\u5305\u300d \u94fe\u63a5\uff1ahttps:\/\/pan.quark.cn\/s\/7d3a8c64d9c3 \u300c\u767e\u5b9d\u7bb1\u300d \u94fe\u63a5\uff1ahttps:\/\/pan.quark.cn\/s\/2bbad26833fc \u300cLTX2.3\u300d\u94fe\u63a5\uff1ahttps:\/\/pan.quark.cn\/s\/17beb2861b6b \u300cFlux2-Klein-Face-Swap\u6362\u8138\u300d\u94fe\u63a5\uff1ahttps:\/\/pan.quark.cn\/s\/54a58a5e0f62 \u300cFlux\u8d44\u6e90\u6574\u5408\u5305\u300d \u94fe\u63a5\uff1ahttps:\/\/pan.quark.cn\/s\/53adf7c46b99 SVIPro\u5355\u56fe\u63d0\u793a\u8bcd\u5168\u81ea\u52a8\u957f\u89c6\u9891\u751f\u6210 \u4e00\u952e\u542f\u52a8\u5305+\u6a21\u578b+\u5de5\u4f5c\u6d41\u94fe\u63a5\uff1ahttps:\/\/pan.quark.cn\/s\/e7e7b5e04a29 \u300cai\u4e16\u754cseedance2\u300d\u94fe\u63a5\uff1ahttps:\/\/pan.quark.cn\/s\/8b83afb0bc63 \u300cComfyUI-seedance2\u5de5\u4f5c\u6d41]:https:\/\/pan.quark.cn\/s\/118540347246 \u300cLTX2.0\u590d\u523bseedance2.0\u300d\u94fe\u63a5\uff1ahttps:\/\/pan.quark.cn\/s\/d58288786f22 \u841d\u535c\u5168\u90e8\u8d44\u6e90\uff1ahttps:\/\/pan.quark.cn\/s\/0a253b3f1d8c \u300csd-webui-aki\u300d\u6574\u5408\u5305\uff1ahttps:\/\/pan.quark.cn\/s\/345008fa69f0 WAN\u6574\u5408\u5305\u5e26\u6a21\u578b\u94fe\u63a5\uff1ahttps:\/\/pan.quark.cn\/s\/b355e6e86853 \u300cAI-Tools\u67e5\u627e\u65b0\u5de5\u5177\u6309\u65e5\u671f\uff0c\u67e5\u627e\u5bf9\u5e94\u5de5\u5177\u6309\u540d\u79f0\u6392\u5e8f\u300d \u94fe\u63a5\uff1ahttps:\/\/pan.quark.cn\/s\/efc430e1996d\"><\/div><\/div><div class=\"lL\" style=\"max-width:100%;width:853px;margin:5px auto;\"><\/div><figcaption><\/figcaption><\/figure>\n\n<figure class=\"wp-block-embed-youtube wp-block-embed is-type-video is-provider-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"lyte-wrapper\" title=\"LTX 2.3 Fine-Tuned Models: Better Consistency, Motion &amp;amp; Prompt Control &mdash;Subtitle &amp;amp; Watermark Removal\" style=\"width:853px;max-width:100%;margin:5px auto;\"><div class=\"lyMe\" id=\"WYL_SQs9X2M9IVA\" itemprop=\"video\" itemscope itemtype=\"https:\/\/schema.org\/VideoObject\"><div><meta itemprop=\"thumbnailUrl\" content=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FSQs9X2M9IVA%2Fhqdefault.jpg\" \/><meta itemprop=\"embedURL\" content=\"https:\/\/www.youtube.com\/embed\/SQs9X2M9IVA\" \/><meta itemprop=\"duration\" content=\"PT24M18S\" \/><meta itemprop=\"uploadDate\" content=\"2026-05-21T13:30:18Z\" \/><\/div><div id=\"lyte_SQs9X2M9IVA\" data-src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FSQs9X2M9IVA%2Fhqdefault.jpg\" class=\"pL\"><div class=\"tC\"><div class=\"tT\" itemprop=\"name\">LTX 2.3 Fine-Tuned Models: Better Consistency, Motion &amp; Prompt Control \u2014Subtitle &amp; Watermark Removal<\/div><\/div><div class=\"play\"><\/div><div class=\"ctrl\"><div class=\"Lctrl\"><\/div><div class=\"Rctrl\"><\/div><\/div><\/div><noscript><a href=\"https:\/\/youtu.be\/SQs9X2M9IVA\" rel=\"nofollow noopener\" target=\"_blank\"><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FSQs9X2M9IVA%2F0.jpg\" alt=\"LTX 2.3 Fine-Tuned Models: Better Consistency, Motion &amp;amp; Prompt Control &mdash;Subtitle &amp;amp; Watermark Removal\" width=\"853\" height=\"460\" \/><br \/>Watch this video on YouTube<\/a><\/noscript><meta itemprop=\"description\" content=\"After Wan stopped releasing open-source versions, LTX 2.3 became the latest and most powerful open-source video generation model available, and the open-source community has produced extensive fine-tunes around it. Back in the Wan 2.1 era, we had FusionX, which delivered better performance and results. Right now, of course, we also have fine-tuned versions of LTX 2.3. In today&#039;s video, I&#039;ll walk through several distinct and worth-trying fine-tuned models based on LTX 2.3 \u2014 Sulphur-2, 10Eros, and ICEdit-Insight. Models\uff1a Sulphur - 2: https:\/\/huggingface.co\/SulphurAI\/Sulphur-2-base\/tree\/main Sulphur - 2 GGUF Version: https:\/\/huggingface.co\/Abiray\/Sulphur-2-base-GGUF\/tree\/main 10Eros: https:\/\/huggingface.co\/TenStrip\/LTX2.3-10Eros\/tree\/main 10Eros-GGUF Version\uff1ahttps:\/\/huggingface.co\/Abiray\/LTX2.3-10Eros-GGUF\/tree\/main ICEdit-Insight https:\/\/huggingface.co\/joyfox\/LTX2.3-ICEdit-Insight\/tree\/main CustomNodes\uff1a 10S-Comfy-nodes: https:\/\/github.com\/TenStrip\/10S-Comfy-nodes KJ Nodes\uff1ahttps:\/\/github.com\/kijai\/ComfyUI-KJNodes Mie Nodes\uff1ahttps:\/\/github.com\/MieMieeeee\/ComfyUI-MieNodes Impact Pack\uff1ahttps:\/\/github.com\/ltdrdata\/ComfyUI-Impact-Pack RES4LYF\uff1ahttps:\/\/github.com\/ClownsharkBatwing\/RES4LYF ComfyUI-LTXVideo\uff1ahttps:\/\/github.com\/Lightricks\/ComfyUI-LTXVide Workflow: https:\/\/drive.google.com\/drive\/folders\/1bdtYD6Ry-wDr_8B_JRrKAAdTSREWygT-?usp=sharing 00:00 \u2014 Effect showcase 01:33 \u2014 Sulphur-2 model intro and walkthrough 14:11 \u2014 10Eros model intro and walkthrough 19:42 \u2014 ICEdit-Insight model intro and walkthrough\"><\/div><\/div><div class=\"lL\" style=\"max-width:100%;width:853px;margin:5px auto;\"><\/div><figcaption><\/figcaption><\/figure>\n\n<figure class=\"wp-block-embed-youtube wp-block-embed is-type-video is-provider-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"lyte-wrapper\" title=\"LTX2.3 Music Video Creator V5.1- ComfyUI - Full Walkthrough\" style=\"width:853px;max-width:100%;margin:5px auto;\"><div class=\"lyMe\" id=\"WYL_LwG-zxY684M\" itemprop=\"video\" itemscope itemtype=\"https:\/\/schema.org\/VideoObject\"><div><meta itemprop=\"thumbnailUrl\" content=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FLwG-zxY684M%2Fhqdefault.jpg\" \/><meta itemprop=\"embedURL\" content=\"https:\/\/www.youtube.com\/embed\/LwG-zxY684M\" \/><meta itemprop=\"duration\" content=\"PT18M26S\" \/><meta itemprop=\"uploadDate\" content=\"2026-05-08T23:43:16Z\" \/><\/div><div id=\"lyte_LwG-zxY684M\" data-src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FLwG-zxY684M%2Fhqdefault.jpg\" class=\"pL\"><div class=\"tC\"><div class=\"tT\" itemprop=\"name\">LTX2.3 Music Video Creator V5.1- ComfyUI - Full Walkthrough<\/div><\/div><div class=\"play\"><\/div><div class=\"ctrl\"><div class=\"Lctrl\"><\/div><div class=\"Rctrl\"><\/div><\/div><\/div><noscript><a href=\"https:\/\/youtu.be\/LwG-zxY684M\" rel=\"nofollow noopener\" target=\"_blank\"><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FLwG-zxY684M%2F0.jpg\" alt=\"LTX2.3 Music Video Creator V5.1- ComfyUI - Full Walkthrough\" width=\"853\" height=\"460\" \/><br \/>Watch this video on YouTube<\/a><\/noscript><meta itemprop=\"description\" content=\"Create automated AI music videos with my full LTX 2.3 workflow for ComfyUI FREE and LOCAL! In this walkthrough, I show how the workflow takes a song, analyzes the timing, creates scene prompts from lyrics, and generates a finished music video using LTX 2.3. The workflow is split into two parts. \ud83c\udfb5 Workflow 1 handles audio upload, beat detection, scene timing, lyrics, style and theme, story idea, subjects and locations, and prompt generation. \ud83c\udfac Workflow 2 handles the actual video generation, including an image-to-video with zImage Turbo and LTX 2.3 workflow, and a text-to-video with LTX, LoRa workflow that both have advanced prompt controls, scene generation, Remake Mode, and final video stitching. \u2728 This workflow is designed to reduce manual setup time while still giving you control over style, characters, camera motion, timing, seeds, LoRas, and final edits. \ud83d\udca1 For the best results, I recommend starting with the default settings first, then experimenting with LoRas, seeds, advanced settings, and Remake Mode as you get more comfortable. \u2699\ufe0f Requirements: ComfyUI LTX 2.3 models Z-Image Turbo model FFmpeg installed for audio stitching My vrgamedevgirl custom nodes Impact Pack custom node for auto-queue llama-cpp-python 16 GB of VRAM min recommended \ud83d\udcac Join my Discord for support, updates, beta features, and to share your work: https:\/\/discord.gg\/ltx23musicvideocreator \u2b07\ufe0f Download my custom nodes and workflows: https:\/\/github.com\/vrgamegirl19\/comfyui-vrgamedevgirl #ComfyUI #LTX #AIvideo #AIMusicVideo #TextToVideo #ImageToVideo #AIWorkflow #GenerativeAI\"><\/div><\/div><div class=\"lL\" style=\"max-width:100%;width:853px;margin:5px auto;\"><\/div><figcaption><\/figcaption><\/figure>\n\n<figure class=\"wp-block-embed-youtube wp-block-embed is-type-video is-provider-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"lyte-wrapper\" title=\"\u7206\u706b\u7684&ldquo;\u65e0\u5ba1\u67e5&rdquo;AI \u89c6\u9891\u6a21\u578b\u6765\u4e86\uff01Sulphur 2 \u672c\u5730\u90e8\u7f72\u5b9e\u6d4b\uff1a8G \u663e\u5b58\u4e5f\u80fd\u8dd1\uff01\u5b8c\u5168\u514d\u8d39\u5f00\u6e90 | \u96f6\u5ea6\u89e3\u8bf4\" style=\"width:853px;max-width:100%;margin:5px auto;\"><div class=\"lyMe\" id=\"WYL_Nms56KhvIQw\" itemprop=\"video\" itemscope itemtype=\"https:\/\/schema.org\/VideoObject\"><div><meta itemprop=\"thumbnailUrl\" content=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FNms56KhvIQw%2Fhqdefault.jpg\" \/><meta itemprop=\"embedURL\" content=\"https:\/\/www.youtube.com\/embed\/Nms56KhvIQw\" \/><meta itemprop=\"duration\" content=\"PT10M11S\" \/><meta itemprop=\"uploadDate\" content=\"2026-05-11T13:12:31Z\" \/><\/div><div id=\"lyte_Nms56KhvIQw\" data-src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FNms56KhvIQw%2Fhqdefault.jpg\" class=\"pL\"><div class=\"tC\"><div class=\"tT\" itemprop=\"name\">\u7206\u706b\u7684\u201c\u65e0\u5ba1\u67e5\u201dAI \u89c6\u9891\u6a21\u578b\u6765\u4e86\uff01Sulphur 2 \u672c\u5730\u90e8\u7f72\u5b9e\u6d4b\uff1a8G \u663e\u5b58\u4e5f\u80fd\u8dd1\uff01\u5b8c\u5168\u514d\u8d39\u5f00\u6e90 | \u96f6\u5ea6\u89e3\u8bf4<\/div><\/div><div class=\"play\"><\/div><div class=\"ctrl\"><div class=\"Lctrl\"><\/div><div class=\"Rctrl\"><\/div><\/div><\/div><noscript><a href=\"https:\/\/youtu.be\/Nms56KhvIQw\" rel=\"nofollow noopener\" target=\"_blank\"><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FNms56KhvIQw%2F0.jpg\" alt=\"\u7206\u706b\u7684&ldquo;\u65e0\u5ba1\u67e5&rdquo;AI \u89c6\u9891\u6a21\u578b\u6765\u4e86\uff01Sulphur 2 \u672c\u5730\u90e8\u7f72\u5b9e\u6d4b\uff1a8G \u663e\u5b58\u4e5f\u80fd\u8dd1\uff01\u5b8c\u5168\u514d\u8d39\u5f00\u6e90 | \u96f6\u5ea6\u89e3\u8bf4\" width=\"853\" height=\"460\" \/><br \/>Watch this video on YouTube<\/a><\/noscript><meta itemprop=\"description\" content=\"\u3010\u66f4\u591a\u8d44\u6e90\u3011\u25b6https:\/\/bittly.cc\/lingdu \u3010\u96f6\u5ea6\u535a\u5ba2\u3011\u25b6https:\/\/www.freedidi.com \u3010\u52a0\u5165\u4f1a\u5458\u3011\u25b6https:\/\/www.youtube.com\/channel\/UCvijahEyGtvMpmMHBu4FS2w\/join \u3010\u9ad8\u7ea7\u4f1a\u5458\u3011\u25b6https:\/\/bittly.cc\/JsLAn ------------------------------------- \ud83d\udd14 Sulphur 2 \u5f00\u6e90\u6a21\u578b\u4e0b\u8f7d\uff1ahttps:\/\/www.freedidi.com\/24142.html \ud83e\udeb6\ud83e\udd47\ud83e\udd48\ud83e\udd49 \u4f1a\u5458\u54a8\u8be2 \/ \u4e00\u5bf9\u4e00\uff08\u9650\u91cf\u5f00\u653e\uff09\uff1a\ud83d\udd17https:\/\/ask.tuio.cc \ud83d\udd30 \u5b89\u5168VPN\u63a8\u8350\uff08\u6700\u9ad8\u4f18\u60e0\u94fe\u63a5\uff09\uff1a 1.SurfsharkVPN\uff1ahttps:\/\/get.surfshark.net\/aff_c?offer_id=390&amp;aff_id=4890 2.ProtonVPN\uff1ahttps:\/\/go.getproton.me\/aff_c?offer_id=26&amp;aff_id=1905&amp;source=youtube 3.ExpressVPN\uff1ahttps:\/\/go.expressvpn.com\/c\/2612933\/1645830\/16063 4.PureVPN\uff1a https:\/\/billing.purevpn.com\/aff.php?aff=44074 \ud83d\ude80VPS \u63a8\u8350\uff1a 1.https:\/\/bandwagonhost.com\/aff.php?aff=58794 \uff08\u642c\u74e6\u5de5 CN2\uff09 \ud83e\udd9e AI \u751f\u6210\u5bf9\u53e3\u578b\u89c6\u9891\uff08\u652f\u6301\u591a\u8bed\u8a00\uff09\uff1ahttps:\/\/bittly.cc\/heygen \ud83d\udcb0\u6bd4\u7279\u5e01\u4ea4\u6613\u5e73\u53f0 1.\u706b\u5e01\u5b98\u7f51 : https:\/\/www.htx.com\/invite\/zh-cn\/1f?invite_code=48tr7 2.\u6b27\u6613\u5b98\u7f51: https:\/\/www.okx.com\/join\/15073927 ======================= \u5176\u5b83\u66f4\u591a\u7cbe\u5f69\u5185\u5bb9\u63a8\u8350\uff1a\ud83d\ude0d https:\/\/youtu.be\/-VrgBEz-EZ4 https:\/\/youtu.be\/bowzVW93pig https:\/\/youtu.be\/uFHtS_kJAUE https:\/\/youtu.be\/dNsxnuAIIdw https:\/\/youtu.be\/q1qCzx3sakA https:\/\/youtu.be\/swafOCGzmow https:\/\/youtu.be\/5eA7-gWH_2o https:\/\/youtu.be\/HAXed-ZYEZQ https:\/\/youtu.be\/nc-EQHtx2W8 https:\/\/youtu.be\/LQwD0x1Z8ak https:\/\/youtu.be\/T4eLUtIRy_M https:\/\/youtu.be\/5XrayNHLWrg https:\/\/youtu.be\/b_NTQEWJLxw https:\/\/youtu.be\/lzb5bJS4XtI https:\/\/youtu.be\/cHIyeTdKB4g https:\/\/youtu.be\/p8KTaBv7nxw https:\/\/youtu.be\/UWelKj2LkIU https:\/\/youtu.be\/JBv-thjmVP8 https:\/\/youtu.be\/QCvF8mCPwdQ https:\/\/youtu.be\/NLO0nx4EN24 https:\/\/youtu.be\/r351ngmDL3Q https:\/\/youtu.be\/IIcAe2AlROE https:\/\/youtu.be\/kCrX3Dm4_cY https:\/\/youtu.be\/UhMim4G79Oo https:\/\/youtu.be\/xzBiF8TJ_t4 https:\/\/youtu.be\/Q-FUmOtysyg https:\/\/youtu.be\/MEZnK8ZOSVw ========================= \u96f6\u5ea6\u7535\u62a5\u7fa4\uff1ahttps:\/\/t.me\/lingdujie\"><\/div><\/div><div class=\"lL\" style=\"max-width:100%;width:853px;margin:5px auto;\"><\/div><figcaption><\/figcaption><\/figure>","protected":false},"excerpt":{"rendered":"<p>Setup All in one LTX Trainer LTX Director LTX IC-LoRA LTX 2.3<\/p>\n","protected":false},"author":1,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"ai_generated_summary":"","footnotes":""},"class_list":["post-8127","page","type-page","status-publish","hentry"],"_links":{"self":[{"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/pages\/8127","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/comments?post=8127"}],"version-history":[{"count":17,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/pages\/8127\/revisions"}],"predecessor-version":[{"id":10069,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/pages\/8127\/revisions\/10069"}],"wp:attachment":[{"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/media?parent=8127"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}