ComfyUI Node
⭐ Star LTXV 2.5 All-in-One (BETA)
All-in-one LTXV 2.5 sampler (official template port). T2V / I2V / I2V+Audio run two passes (half res -> 2x latent upscale -> full res), First/Last-Frame runs a single full-res pass with keyframe guides, Audio Only renders just the soundtrack (64x64, 30 steps). Generated audio is always decoded from the first (high-step) pass. Model+LoRA+CLIP+VAE caching built in. Uses a single ltxv text encoder (LTXV 2.5).
⭐ Star LTXV 2.5 All-in-One (BETA)
- first_frame
- last_frame
- audio
- model_override
- sound_settings
- images
- audio
- frame_rate
◄mode▶️ image_to_video►
◄positive_prompt►
◄negative_promptconsole game, video game, cartoon, childish, ugly►
◄base_modelltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors►
◄clip_1gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors►
◄vaeltx-2.5-video-vae-conv-bf16.safetensors►
◄audio_vaeltx-2.5-audio-vae-bf16.safetensors►
◄upscale_modelltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors►
◄video_sizeHD►
◄ratio1:1►
◄ratio_from_imagetrue►
◄custom_width1024►
◄custom_height1024►
◄frame_rate24►
◄seconds10►
◄seed0►
◄sigma_preset▾►
◄override_audiofalse►
◄lora_1▾►
◄lora_1_strength0.60►
◄lora_2▾►
◄lora_2_strength1.00►
◄lora_3▾►
◄lora_3_strength1.00►
◄custom_sigmas_pass11.0, 0.995833, 0.991667, 0.9875, 0.983333, 0.979167, 0.975, 0.93125, 0.847917, 0.725, 0.522917, 0.28125, 0.0►
◄sigmas_pass20.85, 0.7250, 0.4219, 0.0►
◄cfg1.0►
◄sampler_pass1euler_ancestral►
◄sampler_pass2euler_ancestral►
◄weight_dtype▾►
Category⭐StarNodes/Video
Inputs (35)
| Name | Type | Default | Description |
|---|---|---|---|
| mode | COMBO | ▶️ image_to_video | text_to_video: prompt only. image_to_video: connect first_frame. image_audio_to_video: connect first_frame AND an audio file. first_last_frame_to_video: connect first_frame AND last_frame (single full-res pass with keyframe guides). audio_only: no real video - one 30-step pass at 64x64, only the audio output matters. |
| positive_prompt | STRING | What you want to see. LTXV likes detailed, film-style descriptions with timestamps. | |
| negative_prompt | STRING | console game, video game, cartoon, childish, ugly | What to avoid. Default is the negative prompt from the original workflow. |
| base_model | COMBO | ltx-2.5-22b-distilled-transformer-comfy-int8-convrot.safetensors | LTXV 2.5 A/V checkpoint from models/diffusion_models. Reloaded only when the selection changes. |
| clip_1 | COMBO | gemma4-12b-with-proj-ltx-2.5-comfy-int8-convrot.safetensors | Single LTXV 2.5 text encoder from models/text_encoders (type ltxv). |
| vae | COMBO | ltx-2.5-video-vae-conv-bf16.safetensors | Video VAE from models/vae (e.g. ltx-2.5-video-vae-conv-bf16). |
| audio_vae | COMBO | ltx-2.5-audio-vae-bf16.safetensors | Audio VAE from models/vae (e.g. ltx-2.5-audio-vae-bf16). |
| upscale_model | COMBO | ltx-2.5-latent-spatial-upscaler-x2-bf16-1.0.safetensors | Latent upscaler from models/latent_upscale_models (e.g. ltx-2.5-latent-spatial-upscaler-x2). Used between the passes. |
| video_size | COMBO | HD | HD ~1280px, FHD ~1920px (same tables as the Star LTX Video Settings node), Custom = custom_width/height below. |
| ratio | COMBO | 1:1 | Aspect ratio. Overridden by the input image's ratio when 'ratio_from_image' is enabled and an image is connected. |
| ratio_from_image | BOOLEAN | true | Pick the closest preset ratio to the connected image. Falls back to 'ratio' when no image is connected. |
| custom_width | INT | 102432–8192 | Only used when video_size = Custom. |
| custom_height | INT | 102432–8192 | Only used when video_size = Custom. |
| frame_rate | INT | 241–120 | Frames per second of the output video. |
| seconds | INT | 101–120 | Video length in seconds. Frame count is snapped to 8n+1 (4s @ 25fps = 97 frames). |
| seed | INT | 00–18446744073709550000 | Shared by both sampling passes. |
| sigma_preset | COMBO | Main-pass noise schedule. 8/12/16 = the baked schedules from the original workflow's note node (12 = default, 8 = faster, 16 = finer). 20/30/40/50 = plain sampler steps with the normal scheduler, no custom sigmas. 'custom' uses custom_sigmas_pass1 below. | |
| first_frameopt | IMAGE | Start frame / guide image (image_to_video, image_audio_to_video and first_last_frame_to_video modes). | |
| last_frameopt | IMAGE | Last frame (first_last_frame_to_video mode only). Center-crop resized to the video size and added as the final keyframe. | |
| audioopt | AUDIO | Voice / music track (image_audio_to_video mode). Trimmed to the video length and preserved as-is. | |
| override_audioopt | BOOLEAN | false | text_to_video / image_to_video / first_last_frame_to_video only: when disabled (default), the connected 'audio' input is ignored and the model-generated audio is sent to the audio output. When enabled, the connected 'audio' input is passed straight to the audio output instead. Ignored in image_audio_to_video mode, where the connected audio is always passed through to the output. |
| lora_1opt | COMBO | Optional LoRA stack, applied in order 1 -> 3. | |
| lora_1_strengthopt | FLOAT | 0.60-100–100 | The distilled LoRA in the original workflow ran at 0.6. |
| lora_2opt | COMBO | 1 options: None | |
| lora_2_strengthopt | FLOAT | 1.00-100–100 | — |
| lora_3opt | COMBO | 1 options: None | |
| lora_3_strengthopt | FLOAT | 1.00-100–100 | — |
| custom_sigmas_pass1opt | STRING | 1.0, 0.995833, 0.991667, 0.9875, 0.983333, 0.979167, 0.975, 0.93125, 0.847917, 0.725, 0.522917, 0.28125, 0.0 | Only used when sigma_preset = custom. |
| sigmas_pass2opt | STRING | 0.85, 0.7250, 0.4219, 0.0 | Second-pass (refine) schedule. Default from the workflow. |
| cfgopt | FLOAT | 1.00–100 | Both passes. 1.0 for distilled models, as in the workflow. |
| sampler_pass1opt | COMBO | euler_ancestral | Sampler for pass 1 (half resolution). |
| sampler_pass2opt | COMBO | euler_ancestral | Sampler for pass 2 (full resolution refine). |
| weight_dtypeopt | COMBO | Override base-model dtype. 'default' = as stored. | |
| model_overrideopt | MODEL | Optional external model (e.g. patched with flash/sage attention). When connected, this is used instead of loading 'base_model' from the dropdown, and the LoRA stack below is applied to it directly. | |
| sound_settingsopt | SOUND_SETTINGS | Optional sound processing from a 'Star Video Sound Enricher Option' node - the audio output is cleaned up and enriched with these settings (at least 44.1 kHz, never downsampled) before it leaves the node. |
Outputs (3)
| Name | Type | Description |
|---|---|---|
| images | IMAGE | — |
| audio | AUDIO | — |
| frame_rate | FLOAT | — |