How to Install LTX 2.3 in ComfyUI: Step-by-Step Guide

Hello fellows! This is Dora — and recently I finally cracked the clean install sequence for LTX 2.3. I’d spent two hours on a broken node graph. Right package version, wrong folder. Right folder, missing VAE. Right VAE, wrong ComfyUI build. When it finally rendered a clean 9-frame clip, I immediately wrote down every exact step so I’d never have to do that again.

The blog below explains no vague tips but every command, every folder path, every error I hit and how I fixed it. Let’s get into it.

Prerequisites (ComfyUI version and model storage)

Update ComfyUI before opening an LTX-2.3 workflow. In the Template Library, go to Video and select LTX-2.3 T2V or LTX-2.3 I2V. Follow the workflow’s prompts to download any missing models.

ComfyUI’s LTX-2.3 workflow page lists the required model files and folders, but does not give one universal VRAM minimum or total disk-space requirement for every setup. Check the selected workflow and your available storage before downloading; do not treat the old 0.9.x figures as LTX-2.3 requirements.

Check your ComfyUI build version before proceeding. Open ComfyUI’s terminal and run:

cd ComfyUI
git log --oneline -1

If the commit date is before January 2026, update first:

git pull origin master
pip install -r requirements.txt

Step 1 — Update ComfyUI and open a native workflow

The native LTX-2.3 templates do not require installing ltx-core, ltx-pipelines, or a separate LTX custom-node pack. Update ComfyUI, open Template Library → Video, and choose LTX-2.3 T2V or LTX-2.3 I2V. Follow the prompts to download the workflow’s model files.

If an LTX-2.3 template is missing, ComfyUI may need an update. If a workflow opens with missing core nodes, the latest Nightly may be required; some nodes may not yet be included in the latest stable release. Check ComfyUI’s startup output for core-node import errors.

If you use the ComfyUI portable package (Windows):

# Navigate to your ComfyUI portable folder first
cd C:\ComfyUI_windows_portable
python_embeded\python.exe -m pip install ltx-core ltx-pipelines

If you use a venv or conda environment:

# Activate your environment first
conda activate comfyui   # or: source venv/bin/activate
pip install ltx-core ltx-pipelines

Verify the install worked:

python -c "import ltx_core; print(ltx_core.__version__)"
# Should print: 0.9.2 or higher
python -c "import ltx_pipelines; print('OK')"

If you get ModuleNotFoundError on the verification step, you installed the wrong Python. Re-read the step above and confirm which python executable is actually running ComfyUI.

Step 2 — Download LTX 2.3 Weights

Dev (bf16) vs Distilled model — which to get

I’ve run both through hundreds of generations. Here’s the real tradeoff:

	Dev (bf16)	Distilled
File size	~27 GB	~14 GB
Inference steps	40–50	6–8
Generation time (4090)	~4 min/clip	~45 sec/clip
Motion quality	High	Moderate
Complex scenes	Excellent	Struggles
Prompt iteration	Slow	Fast

My actual workflow: I use distilled for prompt exploration (testing 10–15 prompt variations quickly), then switch to dev for the final output I’ll publish. If you have a 12 GB GPU, start with distilled only.

Where to download (HuggingFace, GitHub monorepo)

All official weights are hosted on HuggingFace by Lightricks:

Download via terminal (faster than browser for large files):

For the native ComfyUI workflow, select T2V or I2V in the Template Library and follow its model-download prompts. If you place files manually, use the filenames and folders listed in the workflow’s current model-storage section below. The LTX-Video 0.9.7 download commands and T5-XXL instructions in older guides are for a different model setup.

Verify file integrity after download (sizes as of March 2026):

Dev bf16: 27.1 GB — if your download is significantly different, re-download
Distilled: 13.8 GB
T5-XXL: 9.8 GB

Install the ComfyUI Custom Node Pack

The model files alone aren’t enough. You need the LTX-specific ComfyUI nodes.

Option A — Via ComfyUI Manager (easiest):

Open ComfyUI in your browser
Click Manager → Install Custom Nodes
Search: ComfyUI-LTX-Video
Install → Restart ComfyUI

Option B — Manual install:

cd ComfyUI/custom_nodes
git clone https://github.com/Lightricks/ComfyUI-LTXVideo
cd ComfyUI-LTXVideo
pip install -r requirements.txt

After installing, fully restart ComfyUI (kill the process — don’t just refresh the browser).

Verify nodes loaded correctly: In the ComfyUI node search (double-click canvas), type LTX. You should see LTXVModelLoader, LTXVSampler, and LTXVScheduler. If these don’t appear, the custom node install failed.

Folder paths and file placement

The current native T2V and I2V workflows list these files:

ComfyUI/└── models/    ├── checkpoints/    │   └── ltx-2.3-22b-dev-fp8.safetensors    ├── loras/    │   ├── ltx_2.3_22b_distilled_1.1_lora_dynamic_fro09_avg_rank_111_bf16.safetensors    │   └── gemma-3-12b-it-abliterated_lora_rank64_bf16.safetensors    ├── text_encoders/    │   └── gemma_3_12B_it_fp4_mixed.safetensors    └── latent_upscale_models/        └── ltx-2.3-spatial-upscaler-x2-1.1.safetensors

These are the filenames and locations shown in ComfyUI’s current LTX-2.3 workflow documentation. Use the selected workflow’s download prompt or compare manually placed files against that list.

Load the official ComfyUI workflow

ComfyUI provides separate native templates for LTX-2.3 text-to-video and image-to-video. Use the matching template and its model list rather than loading a workflow from an older LTX-Video guide.

I2V starter workflow

In Template Library → Video, select LTX-2.3 I2V, or load the official I2V workflow JSON.
Use the template’s default egyptian_queen.png image or provide your own image.
Follow the workflow’s model-download prompts, enter a prompt, and run it.

ComfyUI’s documentation says the I2V template uses the same model set as T2V. Its prompt guidance is to describe actions over time, desired visual details, and any audio or dialogue. For example, “A cyclist turns onto a tree-lined street, then slows at the crossing; the camera follows from behind; distant traffic and a bicycle bell are audible.” This is an illustrative prompt, not a tested result.

Verify Your Install with (Sanity Check)

This guide documents the current official ComfyUI setup path. The T2V and I2V workflows have not been run locally for this article, so no successful generation is claimed.

Common Install Errors and Exact Fixes

Error: ModuleNotFoundError: No module named 'ltx_core'

Cause: Packages installed in wrong Python environment
Fix: Identify which python runs ComfyUI, re-run pip install with that exact executable
Verify: python -c "import ltx_core" must return no error

Error: RuntimeError: size mismatch for transformer.blocks...

Cause: Model file placed in wrong folder (checkpoints/ instead of diffusion_models/)
Fix: Move .safetensors file to ComfyUI/models/diffusion_models/

Error: KeyError: 'LTXVModelLoader' (red nodes in workflow)

Cause: Custom node pack not installed or failed to load
Fix: Check ComfyUI terminal for import errors after restart
     Re-run: pip install -r ComfyUI/custom_nodes/ComfyUI-LTXVideo/requirements.txt

Error: Black video output, no error messages

Cause: VAE node pointing to wrong file (often a generic SD VAE)
Fix: Open VAE loader node, explicitly select ltxvideo_vae_bf16.safetensors

Error: CUDA out of memory on step 1

Cause: VRAM insufficient for chosen model/resolution
Fix: In LTXVModelLoader node, enable "enable_sequential_cpu_offload"
     Or: switch to distilled model and reduce resolution to 512x320

Error: Corrupted or incomplete output video

Cause: T5-XXL text encoder missing or wrong version
Fix: Confirm T5-XXL file is in ComfyUI/models/clip/ and is fp16 version (~9.8 GB)

Keep older LTX checkpoints separate

LTX-2.3, LTX-2.0, and LTX-Video 0.9.x checkpoints are not interchangeable with the native LTX-2.3 workflow. The model-family notes distinguish the LTX-2.3 22B audio-video model from the LTX-Video 0.9.x 2B video-only model, and state that an LTX-2.0 checkpoint does not match the LTX-2.3 contract. Use the checkpoint and supporting files named by the selected LTX-2.3 workflow.

Keep old weights in a separate folder if you need them for archived workflows. Treat 2.3 as a completely fresh install.

AMD GPU (ROCm) — What Actually Works in 2026

Since I see this question constantly: I tested on an RX 7900 XTX (24 GB) running ROCm 6.1 on Ubuntu 22.04. Here’s the real status:

Installation: Works, but requires HSA_OVERRIDE_GFX_VERSION=11.0.0 environment variable set before launching ComfyUI
Generation: Functional but ~2.3× slower than equivalent NVIDIA hardware
Stability: Occasional silent failures mid-generation (no error, just stops)
Windows AMD: Not viable — ROCm on Windows is still too immature

If you’re on AMD/Linux and want to try: the LTX GitHub issues page has a pinned ROCm thread with community-tested launch flags.

FAQ

Why are core nodes missing after I load the workflow? Update ComfyUI. Some core nodes may require the latest Nightly if they are not yet included in the latest stable release. Check the startup output for import errors.

Do I need to install ltx-core or a separate custom-node pack? No. The native LTX-2.3 T2V and I2V templates do not require a separate LTX custom-node pack. ltx-core and ltx-pipelines belong to a separate Python pipeline setup.

Does the native workflow require T5-XXL? No. The current T2V and I2V templates list a Gemma 3 12B text encoder. Follow the filenames in the selected workflow.

What image should I use for I2V? The official I2V template includes egyptian_queen.png; you can also use your own image with that workflow.