LTX 2.3 Model Downloads

Download all LTX 2.3 model files for ComfyUI with direct HuggingFace links — organized by GPU VRAM. Start with taeltx2_3.safetensors (VAE, required for all setups), then choose a checkpoint: FP8 quantized for 16GB VRAM or official full precision for 32GB+.

No GPU? Try LTX 2.3 online — free generation included

Skip the 16GB+ VRAM card and ComfyUI setup. Image-to-video, no install, ~30 seconds per clip.

Try free →

Quick Start: 3 Steps to Get Running

  1. 1.Download taeltx2_3.safetensors (VAE) — required for all workflows
  2. 2.Choose your checkpoint: FP8 v3 (16GB VRAM) or Official Distilled (32GB VRAM)
  3. 3.Place files in ComfyUI folders: VAE → models/vae/, Checkpoint → models/checkpoints/

16GB VRAM — FP8 / MXFP8 Quantized (RTX 40xx+)

FP8 scaled requires RTX 40-series or newer. MXFP8 block-32 is an alternative format for compatible GPUs. NVFP4 (21.7 GB) is the official Blackwell / RTX 50xx path. Use v1.1 FP8 Distilled for fastest generation; use Dev FP8 + LoRA v1.1 if applying LoRA weights.

LTX 2.3 Dev FP8 (Official)

Official FP8
ltx-2.3-22b-dev-fp8.safetensors

Official FP8 from Lightricks. Alternative to Kijai's FP8 dev model.

Official FP8 dev model from Lightricks. 29.1GB, runs on 16GB VRAM.

29.1 GB16GB+ VRAM

LTX 2.3 Distilled FP8 (Official)

Official FP8
ltx-2.3-22b-distilled-fp8.safetensors

Official FP8 distilled from Lightricks. Alternative to Kijai's FP8.

Official FP8 distilled model from Lightricks. 8 steps, CFG=1.

29.5 GB16GB+ VRAM

LTX 2.3 Distilled 1.1 FP8 (Kijai)

16GB Best
ltx-2.3-22b-distilled-1.1_transformer_only_fp8_scaled.safetensors

Best choice for 16GB VRAM. Latest v1.1 FP8 distilled. Requires RTX 40xx+ for fp8 matmuls.

FP8 quantized v1.1 distilled by Kijai. Best for 16GB VRAM. 8 steps, CFG=1.

25.2 GB16GB+ VRAM

LTX 2.3 Distilled 1.1 MXFP8 (Kijai)

🔥 Newv1.1 MXFP8
ltx-2.3-22b-distilled-1.1_transformer_only_mxfp8_block32.safetensors

RTX 30xx workaround — use this when your GPU lacks RTX 40xx-style FP8 matmul support. Same VRAM as fp8_scaled.

MXFP8 block-32 quantized distilled 1.1 by Kijai. Use on RTX 30xx GPUs that cannot run standard FP8 scaled matmul.

~25 GB16GB+ VRAM

LTX 2.3 Dev FP8 (Kijai)

16GB LoRA
ltx-2.3-22b-dev_transformer_only_fp8_input_scaled.safetensors

Use this (not distilled) if you want to apply LoRA weights on 16GB VRAM.

FP8 quantized dev model by Kijai. Runs on 16GB VRAM. Supports LoRA. Place in models/checkpoints/.

25 GB16GB+ VRAM

LTX 2.3 Dev FP8 Scaled (Kijai)

Dev FP8
ltx-2.3-22b-dev_transformer_only_fp8_scaled.safetensors

Alternative FP8 quantization for dev. Use when fp8_input_scaled has compatibility issues.

FP8 scaled dev model by Kijai. 16GB VRAM. Supports LoRA. Place in models/checkpoints/.

~25 GB16GB+ VRAM

LTX 2.3 Dev MXFP8 (Kijai)

Dev MXFP8
ltx-2.3-22b-dev_transformer_only_mxfp8_block32.safetensors

RTX 30xx workaround for dev/quality path — use when standard FP8 matmul is unsupported. Supports LoRA.

MXFP8 block-32 dev model by Kijai. 16GB VRAM. Supports LoRA. Use on RTX 30xx GPUs that cannot run standard FP8 scaled. Place in models/checkpoints/.

~25 GB16GB+ VRAM

LTX 2.3 Dev NVFP4 (Official)

🔥 NewNVFP4 (Blackwell)
ltx-2.3-22b-dev-nvfp4.safetensors

Best size/quality tradeoff on RTX 50xx. On older GPUs it falls back to slow paths — prefer FP8 there.

NVFP4 quantized dev model. Native nvfp4 matmul on Blackwell GPUs (RTX 50xx). 21.7GB.

21.7 GB16GB+ VRAM

LTX 2.3 Distilled LoRA 384 v1.1 (Official)

v1.1 LoRA
ltx-2.3-22b-distilled-lora-384-1.1.safetensors

Latest official LoRA. Pair with dev model for distilled-quality output.

Official distilled LoRA rank-384 v1.1. Use with dev model. Place in models/loras/.

7.61 GB16GB+ VRAM

LTX 2.3 Distilled 1.1 LoRA (Kijai)

16GB LoRA
ltx-2.3-22b-distilled-1.1_lora-dynamic_fro09_avg_rank_111_bf16.safetensors

Pair with the dev FP8 model. Load as LoRA in ComfyUI models/loras/.

Distilled LoRA v1.1 by Kijai. Use with the dev model for distilled-quality output on 16GB VRAM.

2.74 GB16GB+ VRAM

LTX 2.3 Distilled 1.1 LoRA — fro90_ceil72 cond-safe (TenStrip)

🔥 NewI2V LoRA
ltx-2.3-22b-distilled-lora-1.1_fro90_ceil72_condsafe.safetensors

Pair with the dev FP8 model when running image-to-video or other input-conditioned workflows. Use the standard Kijai dynamic-rank LoRA for T2V instead.

Experimental community distilled LoRA v1.1 by TenStrip. The 'cond-safe' variant zeroes cross-attention bridges, adaln/scale-shift tables, gate logits, and prompt scale-shift — making it better suited for I2V and input-conditioned workflows than the standard dynamic LoRA. Place in models/loras/.

662 MB16GB+ VRAM

LTX 2.3 Distilled LoRA Dynamic r105 (Kijai)

Dynamic LoRA
ltx-2.3-22b-distilled-lora-dynamic_fro09_avg_rank_105_bf16.safetensors

v1.0 dynamic LoRA. Use v1.1 dynamic LoRA (rank 111) for latest quality.

Dynamic rank LoRA (avg rank 105) by Kijai. Pair with dev model. Place in models/loras/.

~2.5 GB16GB+ VRAM

24GB VRAM — Official + Sequential Offloading

Enable sequential offloading in ComfyUI settings (Model Offload or Sequential). Uses latest v1.1 official weights.

LTX 2.3 Distilled 1.1 (bf16, 24GB)

24GB
ltx-2.3-22b-distilled-1.1.safetensors

Enable sequential offloading in ComfyUI settings. Uses latest v1.1 official weights.

Official v1.1 distilled model runnable on 24GB with sequential offloading enabled in ComfyUI.

46.1 GB24GB+ VRAM

32GB VRAM — Official Full Precision (BF16)

Official Lightricks checkpoints at full bf16 precision. v1.1 Distilled recommended for most use cases (8 steps, CFG=1). BF16 transformer-only variants from Kijai are also available.

LTX 2.3 Distilled 1.1

v1.1 Latest
ltx-2.3-22b-distilled-1.1.safetensors

For 32GB VRAM: enable Sequential Offloading in ComfyUI settings (file is 46GB). For fastest inference, use the FP8 distilled variant instead.

Official v1.1 distilled model. 8 steps, CFG=1. Latest release from Lightricks. Requires sequential offloading on 32GB — file is 46GB.

46.1 GB32GB+ VRAM

LTX 2.3 Distilled 1.1 BF16 (Kijai)

🔥 Newv1.1 BF16
ltx-2.3-22b-distilled-1.1_transformer_only_bf16.safetensors

Use when FP8 matmul (RTX 40xx+) is unavailable. 44GB — requires 48GB VRAM or sequential offloading on 32GB.

BF16 distilled 1.1 transformer-only by Kijai. For RTX 30xx or GPUs without FP8 support. Requires 48GB+ or sequential offloading on 32GB.

~44 GB32GB+ VRAM

LTX 2.3 Dev

Official
ltx-2.3-22b-dev.safetensors

Best for LoRA training and fine-tuning. 42GB — enable Sequential Offloading on 32GB cards.

Full BF16 dev model. Flexible and trainable. 42GB — requires 48GB VRAM or sequential offloading on 32GB.

~42 GB32GB+ VRAM

LTX 2.3 Dev BF16 (Kijai)

Dev BF16
ltx-2.3-22b-dev_transformer_only_bf16.safetensors

Use when training LoRA with full BF16 precision. 44GB — enable Sequential Offloading on 32GB cards.

BF16 dev transformer-only by Kijai. 44GB — requires 48GB VRAM or sequential offloading on 32GB. Place in models/checkpoints/.

~44 GB32GB+ VRAM

LTX 2.3 Distilled LoRA 384 v1.1 (Official)

v1.1 LoRA
ltx-2.3-22b-distilled-lora-384-1.1.safetensors

Latest official LoRA. Pair with dev model for distilled-quality output.

Official distilled LoRA rank-384 v1.1. Use with dev model. Place in models/loras/.

7.61 GB16GB+ VRAM

LTX 2.3 Distilled LoRA 384 (Official)

v1.0 LoRA
ltx-2.3-22b-distilled-lora-384.safetensors

Previous v1.0 LoRA. Use v1.1 LoRA for latest quality.

Official distilled LoRA rank-384 v1.0. Pair with dev model. Place in models/loras/.

~7.6 GB16GB+ VRAM

IC-LoRA Family — Control, HDR, LipDub (Official Lightricks)

IC-LoRAs (In-Context LoRAs) attach to the dev model for structural control (pose/depth/edges), motion trajectories, 16-bit HDR generation, and lip-dub. Use the official IC-LoRA ComfyUI workflow with LTXICLoRALoaderModelOnly + LTXAddVideoICLoRAGuide at Reference Downscale Factor 0.5. Place in models/loras/. — See Guide step 7 for the full how-to.

IC-LoRA Union Control

🔥 NewIC-LoRA Union
ltx-2.3-22b-ic-lora-union-control-ref0.5.safetensors

Use the official IC-LoRA ComfyUI workflow with LTXICLoRALoaderModelOnly + LTXAddVideoICLoRAGuide (Reference Downscale Factor 0.5).

Unified Canny + Depth control IC-LoRA. One LoRA covering edge and depth conditioning. Place in models/loras/.

654 MB16GB+ VRAM

IC-LoRA Motion Track Control

🔥 NewIC-LoRA Motion
ltx-2.3-22b-ic-lora-motion-track-control-ref0.5.safetensors

For V2V motion transfer. Pair with the official IC-LoRA workflow at Reference Downscale Factor 0.5.

Motion-track IC-LoRA. Drive motion with sparse point trajectories (SpatialTrackerV2 or hand-drawn). Place in models/loras/.

327 MB16GB+ VRAM

IC-LoRA HDR

🔥 NewIC-LoRA HDR
ltx-2.3-22b-ic-lora-hdr-0.9.safetensors

Required: also download the HDR scene-emb file. Enables HDR text/image-to-video and SDR-to-HDR conversion.

16-bit HDR generation IC-LoRA. SDR→HDR via LogC3 transform. Place in models/loras/.

327 MB16GB+ VRAM

IC-LoRA HDR Scene Embeddings

🔥 New
ltx-2.3-22b-ic-lora-hdr-scene-emb.safetensors

Required when using the HDR IC-LoRA.

Companion scene-embeddings file for the HDR IC-LoRA. Place in models/loras/.

12.6 MB4GB+ VRAM

IC-LoRA LipDub

🔥 NewIC-LoRA LipDub
ltx-2.3-22b-ic-lora-lipdub-0.9.safetensors

Use for video lip-sync / dubbing. Pair with the audio VAE.

Lip-dubbing IC-LoRA based on JustDubIt research. Joint audio-visual diffusion for dubbing. Place in models/loras/.

2.47 GB16GB+ VRAM

Text Encoders — Gemma 3 12B IT (Required)

Every LTX 2.3 ComfyUI workflow needs a Gemma 3 12B text encoder. Use FP4 mixed (9.5 GB) on 16/24 GB cards; the full BF16 file is for 32 GB+. Workflows reference the BF16 file as 'comfy_gemma_3_12B_it.safetensors' — rename after download. Place in models/text_encoders/. — See Guide step 5.

Gemma 3 12B IT FP4 Mixed (Text Encoder)

🔥 NewRequired (16/24 GB)
gemma_3_12B_it_fp4_mixed.safetensors

Use this on 16/24 GB VRAM cards. Full BF16 Gemma OOMs alongside the transformer.

FP4-mixed Gemma 3 12B IT text encoder (~90% FP4 layers). Required for 16-24 GB ComfyUI workflows. Place in models/text_encoders/.

9.5 GB8GB+ VRAM

Gemma 3 12B IT FP8 Scaled (Text Encoder)

🔥 New
gemma_3_12B_it_fp8_scaled.safetensors

Alternative for 16-24 GB cards. Slightly higher quality than FP4 at 3.7 GB more.

FP8-scaled Gemma 3 12B IT text encoder. Alternative to FP4 with marginally higher precision. Place in models/text_encoders/.

13.2 GB16GB+ VRAM

Gemma 3 12B IT BF16 (Text Encoder)

🔥 NewRequired (32 GB+)
comfy_gemma_3_12B_it.safetensors

Rename to comfy_gemma_3_12B_it.safetensors after downloading. Use on 32 GB+ cards for highest text-encoder quality.

Full BF16 Gemma 3 12B IT text encoder (HF filename gemma_3_12B_it.safetensors — workflows reference it as comfy_gemma_3_12B_it.safetensors, rename after download). Place in models/text_encoders/.

24.4 GB32GB+ VRAM

LTX 2.3 Text Projection (Kijai)

ltx-2.3_text_projection_bf16.safetensors

Required for workflows using separate text encoder components.

Text projection BF16 component by Kijai. Place in models/text_encoders/.

~0.5 GB2GB+ VRAM

Previous Versions — v1.0 Models

v1.0 variants superseded by v1.1. Listed for reference or compatibility with existing workflows.

LTX 2.3 Distilled

ltx-2.3-22b-distilled.safetensors

Previous version. Use v1.1 Distilled for latest quality.

v1.0 distilled version. 8 steps, CFG=1. Superseded by v1.1.

~42 GB32GB+ VRAM

LTX 2.3 Distilled FP8 v3 (Kijai)

ltx-2.3-22b-distilled_transformer_only_fp8_input_scaled_v3.safetensors

Previous version. Use v1.1 FP8 for latest quality.

FP8 distilled v3 by Kijai. Previous version, superseded by v1.1 FP8.

25 GB16GB+ VRAM

LTX 2.3 Distilled FP8 v1 (Kijai)

ltx-2.3-22b-distilled_transformer_only_fp8_input_scaled.safetensors

Previous version. Use FP8 v3 or v1.1 FP8 for better quality.

FP8 distilled v1 by Kijai. Earliest FP8 release, superseded by v3.

~25 GB16GB+ VRAM

LTX 2.3 Distilled FP8 v2 (Kijai)

ltx-2.3-22b-distilled_transformer_only_fp8_input_scaled_v2.safetensors

Previous version. Use FP8 v3 or v1.1 FP8 for better quality.

FP8 distilled v2 by Kijai. Superseded by v3.

~25 GB16GB+ VRAM

LTX 2.3 Distilled FP8 Scaled (Kijai)

ltx-2.3-22b-distilled_transformer_only_fp8_scaled.safetensors

Alternative FP8 quantization. Use v1.1 FP8 for latest quality.

FP8 scaled distilled by Kijai. Alternative FP8 quantization method.

~25 GB16GB+ VRAM

LTX 2.3 Distilled MXFP8 (Kijai)

ltx-2.3-22b-distilled_transformer_only_mxfp8_block32.safetensors

Previous v1.0 MXFP8 (RTX 30xx). Use v1.1 MXFP8 for latest quality.

MXFP8 block-32 distilled v1.0 by Kijai. RTX 30xx workaround — use when standard FP8 scaled is unsupported.

~25 GB16GB+ VRAM

LTX 2.3 Distilled BF16 (Kijai)

ltx-2.3-22b-distilled_transformer_only_bf16.safetensors

Previous v1.0 BF16 distilled. Use v1.1 BF16 for latest quality. Enable Sequential Offloading on 32GB.

BF16 distilled v1.0 transformer-only by Kijai. 44GB — requires 48GB VRAM or sequential offloading on 32GB.

~44 GB32GB+ VRAM

LTX 2.3 Dev FP8 Input Scaled (dash filename, Kijai)

ltx-2-3-22b-dev_transformer_only_fp8_input_scaled.safetensors

Prefer the dotted-filename version unless a specific workflow references this exact name.

FP8 input-scaled dev model under the alternate `ltx-2-3-…` (dash) filename. Same weights, different name.

25 GB16GB+ VRAM

Required & Optional Components

taeltx2_3.safetensors (VAE) is required for all setups. Audio VAE enables audio-conditioned workflows. Upscalers are optional — place in models/latent_upscale_models/.

LTX 2.3 VAE

Required
taeltx2_3.safetensors

Required for all setups. Download this first regardless of your VRAM.

VAE by Kijai. Required for all ComfyUI workflows. Place in models/vae/.

~0.5 GB2GB+ VRAM

LTX 2.3 Audio VAE (Kijai)

🔥 NewNew: Audio
LTX23_audio_vae_bf16.safetensors

Required for audio-conditioned workflows. Unlocks new audio-to-video capability.

Audio VAE for audio-conditioned video generation. Place in models/vae/. Enables audio-to-video workflows.

~1 GB2GB+ VRAM

LTX 2.3 Video VAE (Kijai)

LTX23_video_vae_bf16.safetensors

Alternative VAE option. Use in workflows that require the separate BF16 VAE component.

Standalone video VAE BF16 by Kijai. Alternative to taeltx2_3. Place in models/vae/.

~1 GB2GB+ VRAM

Spatial Upscaler x2 v1.1

v1.1
ltx-2.3-spatial-upscaler-x2-1.1.safetensors

Updated upscaler. Prefer over v1.0 for better upscaling quality.

Updated spatial upscaler x2 v1.1. Place in models/latent_upscale_models/.

~1 GB4GB+ VRAM

Spatial Upscaler x2

ltx-2.3-spatial-upscaler-x2-1.0.safetensors

Optional. Use in a two-stage pipeline to upscale output resolution after generation.

Spatial upscaler x2 for two-stage pipelines. Place in models/latent_upscale_models/.

~1 GB4GB+ VRAM

Spatial Upscaler x1.5

ltx-2.3-spatial-upscaler-x1.5-1.0.safetensors

Use when x2 upscale is too aggressive. Gentler upscaling option.

Spatial upscaler x1.5 for two-stage pipelines. Place in models/latent_upscale_models/.

1.09 GB4GB+ VRAM

Temporal Upscaler x2

ltx-2.3-temporal-upscaler-x2-1.0.safetensors

Optional. Use to double frame count (e.g. 65→129 frames) for smoother motion.

Temporal upscaler x2 for frame interpolation. Place in models/latent_upscale_models/.

262 MB4GB+ VRAM
Official Guide

How to Choose the Right Model

Distilled vs Dev

Distilled — 8 steps, CFG=1. Recommended for most users. Fastest generation, high quality. Cannot be fine-tuned.

Dev — Full model. Use only if you need LoRA training or fine-tuning. Slower, more flexible.

FP8 vs MXFP8 vs BF16

FP8 scaled (Kijai) — Standard 8-bit float. Runs on 16GB. Requires RTX 40xx+. Slight quality trade-off.

MXFP8 block-32 (Kijai) — Alternative FP8 format. Try if standard FP8 causes errors on your GPU.

BF16 (Official / Kijai) — Full precision. Best quality. Needs 32GB+ VRAM.

Quick decision

  • 16GB, RTX 40xx+: Distilled 1.1 FP8 scaled
  • 16GB, FP8 issues: Distilled 1.1 MXFP8 block-32
  • 16GB + LoRA: Dev FP8 + LoRA 1.1
  • 24GB: Official Distilled 1.1 + offloading
  • 32GB+: Official Distilled 1.1 (recommended)

Always required

taeltx2_3.safetensors (VAE) — place in models/vae/. Every workflow needs this regardless of which checkpoint you use. For audio-to-video workflows, also download LTX23_audio_vae_bf16.safetensors.

Official ComfyUI-LTXVideo README →

Setup Guide →

How to install and configure LTX 2.3 with ComfyUI

Workflow Templates →

Download ComfyUI workflow JSON for LTX 2.3