File size: 3,034 Bytes
f638bf2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
[
  {
    "name": "Text to image (SD-Turbo, 1 step)",
    "file": "txt2img-sd-turbo.json",
    "note": "The smallest model: draws in one step."
  },
  {
    "name": "Text to image (Stable Diffusion 1.5, real guidance)",
    "file": "txt2img-sd15.json",
    "note": "20 steps with a negative prompt; downloads 4.4 GB the first time."
  },
  {
    "name": "Image to image",
    "file": "img2img.json",
    "note": "Upload a picture, pick it in Load Image, then Queue."
  },
  {
    "name": "Hires fix (draw small, upscale the latent, refine)",
    "file": "hires-fix.json",
    "note": "SD-Turbo at 256 px, then 512 px."
  },
  {
    "name": "Inpainting (redraw the masked coat)",
    "file": "inpaint.json",
    "note": "The coat in the example picture is already masked (transparent), so Queue redraws it from the prompt. For your own: upload a picture, right-click Load Image > Open in Mask Editor, paint, Queue."
  },
  {
    "name": "FLUX.2 Klein text to image",
    "file": "flux2-klein.json",
    "note": "4B model, 4 steps; downloads 4.6 GB the first time. Needs a GPU with about 4 GB free."
  },
  {
    "name": "FLUX 3 Klein text to image",
    "file": "flux3-klein.json",
    "note": "5.6B model, 16 steps, wants a full scene caption; downloads 6 GB the first time."
  },
  {
    "name": "FLUX.2 Klein picture edit",
    "file": "flux2-klein-edit.json",
    "note": "Edits the picture you load, following the prompt."
  },
  {
    "name": "Text to video (LTX-Video 2B)",
    "file": "ltx-video-t2v.json",
    "note": "A one-second clip (25 frames at 24 fps) from a prompt, saved as an animated WEBP. The first run downloads about 8 GB of weights."
  },
  {
    "name": "Picture to video (LTX-Video 2B)",
    "file": "ltx-video-i2v.json",
    "note": "The example picture as the first frame of a one-second clip; the prompt says what happens next. The first run downloads about 8 GB of weights."
  },
  {
    "name": "LLM writes the prompt, SD-Turbo draws it",
    "file": "llm-prompt.json",
    "note": "A small local language model expands your idea into a prompt."
  },
  {
    "name": "Upscale a picture with a model",
    "file": "upscale.json",
    "note": "ESPCN x3, a few hundred KB."
  },
  {
    "name": "Remove the background",
    "file": "remove-background.json",
    "note": "RMBG 1.4: the mask, and the cut-out on green."
  },
  {
    "name": "Depth map",
    "file": "depth.json",
    "note": "Depth Anything V2 small."
  },
  {
    "name": "Pose and object detection",
    "file": "detect.json",
    "note": "Boxes around what is in the picture, the list of what was found, and the body skeleton."
  },
  {
    "name": "Speak and listen (Kokoro, then Parakeet)",
    "file": "speech.json",
    "note": "Text to speech (Kokoro), then the speech back to text (Parakeet)."
  },
  {
    "name": "Image to 3D model (TRELLIS.2)",
    "file": "image-to-3d.json",
    "note": "Upload a picture of one object, pick it in Load Image, Queue; right-click Save 3D Model > View 3D model. About 10 GB the first time."
  }
]