ComfyUI Node
toobusy Z-Image Turbo
A ComfyUI node in toobusy/Make with 42 inputs and 10 outputs.
toobusy Z-Image Turbo
- image
- model_override
- clip_override
- vae_override
- positive_override
- negative_override
- latent_override
- zit_control
- image
- latent
- width
- height
- model
- model_clean
- clip
- vae
- positive
- negative
◄model_nameZIT/zImage_turbo.safetensors►
◄clip_nameZIT/zImage_textEncoder.safetensors►
◄vae_nameFLUX1/ae.safetensors►
◄positive►
◄negative►
◄ratio_preset2:3►
◄megapixels1.00►
◄divisible_by32►
◄batch_size1►
◄seed1►
◄steps8►
◄cfg1.0►
◄sampler_nameres_multistep►
◄schedulersimple►
◄denoise1.00►
◄aura_shift3.0►
◄lora_slots0►
◄lora_1_enablefalse►
◄lora_1_nameLora/ZIT/ZIT_Neobabae_v1.safetensors►
◄lora_1_strength1.00►
◄lora_2_enablefalse►
◄lora_2_nameNone►
◄lora_2_strength1.00►
◄lora_3_enablefalse►
◄lora_3_nameNone►
◄lora_3_strength1.00►
◄lora_4_enablefalse►
◄lora_4_nameNone►
◄lora_4_strength1.00►
◄lora_5_enablefalse►
◄lora_5_nameNone►
◄lora_5_strength1.00►
◄width0►
◄height0►
Categorytoobusy/Make
Inputs (42)
| Name | Type | Default | Description |
|---|---|---|---|
| model_name | COMBO | ZIT/zImage_turbo.safetensors | 1 options: ZIT/zImage_turbo.safetensors |
| clip_name | COMBO | ZIT/zImage_textEncoder.safetensors | 1 options: ZIT/zImage_textEncoder.safetensors |
| vae_name | COMBO | FLUX1/ae.safetensors | 1 options: FLUX1/ae.safetensors |
| positive | STRING | — | |
| negative | STRING | — | |
| ratio_preset | COMBO | 2:3 | 9 options: 1:1, 16:9, 9:16, 4:3, 3:4, 3:2, +3 |
| megapixels | FLOAT | 1.000.1–4 | — |
| divisible_by | INT | 328–128 | — |
| batch_size | INT | 11–64 | — |
| seed | INT | 10–18446744073709550000 | — |
| steps | INT | 81–100 | — |
| cfg | FLOAT | 1.00–30 | — |
| sampler_name | COMBO | res_multistep | 44 options: euler, euler_cfg_pp, euler_ancestral, euler_ancestral_cfg_pp, heun, heunpp2, +38 |
| scheduler | COMBO | simple | 9 options: simple, sgm_uniform, karras, exponential, ddim_uniform, beta, +3 |
| denoise | FLOAT | 1.000–1 | — |
| aura_shift | FLOAT | 3.00–20 | — |
| lora_slots | INT | 00–5 | — |
| lora_1_enable | BOOLEAN | false | — |
| lora_1_name | COMBO | Lora/ZIT/ZIT_Neobabae_v1.safetensors | 2 options: None, Lora/ZIT/ZIT_Neobabae_v1.safetensors |
| lora_1_strength | FLOAT | 1.00-10–10 | — |
| lora_2_enable | BOOLEAN | false | — |
| lora_2_name | COMBO | None | 2 options: None, Lora/ZIT/ZIT_Neobabae_v1.safetensors |
| lora_2_strength | FLOAT | 1.00-10–10 | — |
| lora_3_enable | BOOLEAN | false | — |
| lora_3_name | COMBO | None | 2 options: None, Lora/ZIT/ZIT_Neobabae_v1.safetensors |
| lora_3_strength | FLOAT | 1.00-10–10 | — |
| lora_4_enable | BOOLEAN | false | — |
| lora_4_name | COMBO | None | 2 options: None, Lora/ZIT/ZIT_Neobabae_v1.safetensors |
| lora_4_strength | FLOAT | 1.00-10–10 | — |
| lora_5_enable | BOOLEAN | false | — |
| lora_5_name | COMBO | None | 2 options: None, Lora/ZIT/ZIT_Neobabae_v1.safetensors |
| lora_5_strength | FLOAT | 1.00-10–10 | — |
| imageopt | IMAGE | Connect an image to switch this node to img2img. The image is VAE-encoded as the starting latent and 'denoise' becomes the change strength (lower = closer to the source). Leave unconnected for plain text-to-image. | |
| widthopt | INT | 00–8192 | 0 = use ratio_preset + megapixels. Set width AND height > 0 to enter the resolution directly (rounded to divisible_by). In img2img this scales the source image to the given size. |
| heightopt | INT | 00–8192 | 0 = use ratio_preset + megapixels. Set width AND height > 0 to enter the resolution directly (rounded to divisible_by). In img2img this scales the source image to the given size. |
| model_overrideopt | MODEL | — | |
| clip_overrideopt | CLIP | — | |
| vae_overrideopt | VAE | — | |
| positive_overrideopt | CONDITIONING | External positive conditioning. Connected = the positive prompt text is ignored (and the internal CLIP is not even loaded when both overrides are connected and no LoRA is active). | |
| negative_overrideopt | CONDITIONING | External negative conditioning. Connected = the negative prompt text is ignored. | |
| latent_overrideopt | LATENT | External starting latent (e.g. from toobusy Hires Upscale). Connected = wins over the image input and the empty latent; 'denoise' is the change strength. Width/height follow the latent. | |
| zit_controlopt | ZIT_CONTROL | Connect a 'toobusy ZIT ControlNet' module to apply depth/canny/pose Fun-ControlNet-Union model patches. Leave unconnected for plain generation. |
Outputs (10)
| Name | Type | Description |
|---|---|---|
| image | IMAGE | — |
| latent | LATENT | — |
| width | INT | — |
| height | INT | — |
| model | MODEL | — |
| model_clean | MODEL | — |
| clip | CLIP | — |
| vae | VAE | — |
| positive | CONDITIONING | — |
| negative | CONDITIONING | — |