CivArchive

    LTX-2 19B — COMPLETE COMMUNITY TOOLKIT

    Everything you need to start creating with LTX-2 19B in ComfyUI, gathered in one community-maintained Civitai hub: base models, GGUF and FP8 variants, text encoders, VAEs, LoRAs, upscalers, ControlNet resources and ready-to-use workflows.

    IMPORTANT

    LTX-2 was created by Lightricks. I am not the original creator of the base model. This page mirrors and organizes useful public files from the official Lightricks ecosystem while also hosting my own workflows and community resources.

    Official source: https://huggingface.co/Lightricks

    START HERE — CHOOSE ONE MAIN MODEL

    You do not need to download every file on this page.

    • GGUF Q4_M — Smallest GGUF option for constrained hardware

    • GGUF Q6_K — Good balance between size and fidelity

    • GGUF Q8_0 — Highest-precision GGUF option, with the largest footprint

    • FP8 Distilled — Faster generation and rapid iteration

    • FP8 Dev — Full model behavior and maximum workflow flexibility

    The distilled workflow is the easiest starting point when speed matters. Choose the full FP8 or higher-precision GGUF variants when final quality and flexibility matter more.

    INCLUDED RESOURCES

    • LTX-2 19B GGUF models: Q4_M, Q6_K and Q8_0

    • Standard and distilled FP8 models

    • Gemma text encoders

    • Video and audio VAE components

    • Distilled, detail-enhancement and camera-control LoRAs

    • Spatial and temporal upscalers

    • Canny ControlNet

    • EKKIVOK text-to-video and image-to-video workflows

    COMFYUI QUICK SETUP

    1. Update ComfyUI before loading an LTX-2 workflow.

    2. For native workflows, open Template Library → Video → LTX-2.

    3. For GGUF models, install ComfyUI-GGUF and use Unet Loader (GGUF). Place .gguf files in ComfyUI/models/unet/; recent versions can also expose ComfyUI/models/diffusion_models/.

    4. Place LoRAs in ComfyUI/models/loras/ and text encoders in ComfyUI/models/text_encoders/.

    5. Load the included workflow, install missing custom nodes through ComfyUI Manager, then restart ComfyUI.

    6. Select the downloaded filenames in the corresponding loader nodes. A downloaded filename may differ from the name saved inside an older workflow, so relink it once if ComfyUI shows a missing model.

    TROUBLESHOOTING

    • GGUF model missing: verify that you are using Unet Loader (GGUF), refresh the model list and restart ComfyUI.

    • Red or missing nodes: use Install Missing Custom Nodes in ComfyUI Manager, then restart.

    • Workflow asks for a differently named file: select the equivalent file manually in the loader node; you do not need to download the entire repository.

    • Out of memory: begin with a smaller GGUF quantization or the distilled workflow, reduce resolution and frame count, and avoid loading unused models simultaneously.

    • Need help: include your GPU, VRAM, ComfyUI version, workflow name and exact error message in your comment.

    CREDITS AND DOCUMENTATION

    • Original LTX project: https://huggingface.co/Lightricks

    • Official LTX documentation: https://docs.ltx.video/open-source-model/integration-tools/comfy-ui

    • Official ComfyUI LTX-2 guide: https://docs.comfy.org/tutorials/video/ltx/ltx-2

    • ComfyUI-GGUF: https://github.com/city96/ComfyUI-GGUF

    If this hub or one of my workflows helps you, sharing a result and leaving a review is the best way to support continued updates.

    Description

    FAQ

    Comments (15)

    StarWJan 16, 2026
    CivitAI

    CLIPTextEncode

    mat1 and mat2 shapes cannot be multiplied (1024x9984 and 188160x3840)

    keep getting this error. everything is loaded exactly like you instructed .....

    EKKIVOK
    Author
    Jan 16, 2026

    wrong encoder. Use the one i uploaded on this page :)

    HavemoneyJan 16, 2026
    CivitAI

    On the LTX reddit they say an updated VAE is available. Which one is posted here?

    EKKIVOK
    Author
    Jan 16, 2026

    ithink its the smallest one. but what kind of vae you're talking about video vae or audio vae ?
    cause there is two types of vae for ltx @Havemoney

    CernerJan 16, 2026· 3 reactions
    CivitAI

    Guide from the LTX team on how to train your own lora, they specialized their own free training system: https://www.youtube.com/watch?v=sL-T6dsO0v4

    EKKIVOK
    Author
    Jan 16, 2026

    Thanks bro!! that will be usefull i didn't found one with ostris toolkit <3

    chrmaroudas625Jan 17, 2026· 1 reaction
    CivitAI

    By the gods at last all in one place instead of hugging face repos with missing files, GGUFs repos with missing workflows, and "workflows" with a million custom nodes like we're in arch linux

    Thank you man

    EKKIVOK
    Author
    Jan 17, 2026

    NO problem upload still in progress :) you can follow the status in the main description :)

    amgjustsaiyan595Jan 17, 2026
    CivitAI

    Thank you for arrenging every thing in one place . i have rtx 5080 -16GB . will this work for me? gonna test anyway ;) .

    EKKIVOK
    Author
    Jan 17, 2026

    Oh sure it will lol :D

    FemBroJan 18, 2026

    Yeah you can use the Quantized GGUF's by Unsloth or City96

    amgjustsaiyan595Feb 20, 2026

    @EKKIVOK I AM GETTING ERROR - LTXVGemmaCLIPModelLoader

    No files matching pattern 'tokenizer.model' found under E:\Comfy UI\ComfyUI_windows_portable_nvidia\ComfyUI_windows_portable\ComfyUI\models

    amgjustsaiyan595Feb 20, 2026

    that solved after downloading tokenizer files .now getting error - LTXVGemmaCLIPModelLoader

    unsupported operand type(s) for //: 'int' and 'list'

    fleetwood_70961Jan 17, 2026
    CivitAI

    Has anyone else had unwanted, phantom, gibberish text appear in lower third in the LTX-2 i2v? I cannot make it disappear.

    EKKIVOK
    Author
    Jan 17, 2026

    fleetwood_70961 yeah i had that problem and i get this disapeared by adding in negative prompt : Text, letters, Watermarks.