Everything You Need to Start Creating with LTX 2 - 12 GB VRAM
Added 2026-02-04 22:16:10 +0000 UTCHi guys!
Thanks so much for watching my latest video on LTX 2. This model is incredible, and I want to make sure you have the best experience using it. Below, you will find all the models, links, and workflows you need to replicate what I did in the video. Thank you for making this content possible with your support!
Whether youβre diving in with the manual setup or using the one-click (paid tier perk π), I hope you have an absolute blast creating with LTX-2. You guys are the reason I keep pushing β thank you from the bottom of my heart. β€οΈ
Link to my installer:
https://www.patreon.com/posts/ltx-2-12gb-vram-149395192
Trouble prompting your actions? I made this a Guide a while back when Wan 2.1 came out. A 69 Page Guide for my paid members. Most of it also applies to LTX (except for the links to the models ofc) :
https://www.patreon.com/posts/light-camera-wan-130532930
π Python Requirements
Python Version: 3.12
https://www.python.org/ftp/python/3.12.10/python-3.12.10-amd64.exe
π¦ Git Repositories You Need
Main Repository (ComfyUI):
https://github.com/comfyanonymous/ComfyUI
or full installer
https://www.comfy.org/
Additional Repositories
ComfyUI-Manager
https://github.com/Comfy-Org/ComfyUI-Manager
ComfyUI-GGUF
https://github.com/city96/ComfyUI-GGUF
ComfyUI-KJNodes
https://github.com/kijai/ComfyUI-KJNodes
cg-use-everywhere
https://github.com/chrisgoringe/cg-use-everywhere
rgthree-comfy
https://github.com/rgthree/rgthree-comfy
ComfyUI_essentials
https://github.com/cubiq/ComfyUI_essentials
ComfyUI-MelBandRoFormer
https://github.com/kijai/ComfyUI-MelBandRoFormer
comfy-audio-duration
https://github.com/scofano/comfy-audio-duration
π¨ Models You Need
πΌ Diffusion Models
(Tip: Place in your models\diffusion_models folder)
π₯ MelBandRoFormer_comfy
VRAM Options (pick one that fits your VRAM):
π₯ LTX-2-GGUF
VRAM Options (pick one that fits your VRAM):
π Text Encoders
(Tip: Place in your models\text_encoders folder)
π₯ gemma_3_12B_it_fp4_mixed
π VAEs
(Tip: Place in your models\vae folder)
π₯ LTX2_video_vae_bf16
π₯ LTX2_audio_vae_bf16
π VAE Approx
(Tip: Place in your models\vae_approx folder)
π₯ taeltx_2
π§ LoRAs
(Tip: Place in your models\loras folder)
π₯ ltx-2-19b-distilled-lora-resized_dynamic_fro095_avg_rank_242_bf16
VRAM Options (pick one that fits your VRAM):
π₯ LTX-2-19b-IC-LoRA-Detailer
π₯ ltx-2-19b-lora-camera-control-static
π CLIP Models
(Tip: Place in your models\clip folder)
π₯ ltx-2-19b-embeddings_connector_dev_bf16
π Upscale Models
(Tip: Place in your models\latent_upscale_models folder)
π₯ ltx-2-spatial-upscaler-x2-1.0