Skip to content

Model Checkpoint Comparison

Video Lecture

Section Video Links
Model Checkpoint Comparison Model Checkpoint Comparison Model Checkpoint Comparison
Video Timings 00:00 Model Comparison Introduction
00:30 Model Information and Settings
01:00 Overview of Models
01:45 Downloading and Installing Models
02:15 Comfy UI Workflow Setup
02:45 Demonstration Prompt
03:00 Stable Diffusion 1.5 Test
03:45 Stable Diffusion 2.1 Test
04:30 Stable Diffusion XL Test
05:30 Stable Diffusion 3.5 Test
07:00 Flux Schnell Test
08:30 Flux Dev Test
09:30 Dream Shaper 8 Test
10:15 Absolute Reality Test
10:45 Dream Shaper XL Test
11:45 Conclusion and Experimentation Advice
12:15 Flux Schnell Retest

Description

In this video, we will do a quick overview of popular image generation models such as SD 1.5, SD 2.1, SDXL, SD 3.5, Flux Dev, Flux Schnell, DreamShaper 8, Dreamshaper XL and AbsoluteReality.

There is no one model that is best at everything, they are all trained on different data sets, and there are many models available. But you can experiment to find a compromise between acceptable speed and quality.

Stable Diffusion 1.5

Stable Diffusion 2.1

  • Download : SD2.1 512x512 (huggingface) | SD2.1 768x768 (huggingface)
  • Training Image Resolution: 512×512 or 768×768
  • Ideal KSampler Input Image Resolution: 512×512 or 768×768
  • Other Ksampler Settings: Steps 20, CFG 4, Euler, Normal
  • Approx. VRAM Required: ~6 GB
  • Realism: Improved over SD1.5; clearer details.
  • People: Better anatomy and faces, but still imperfect.
  • Landscapes: Enhanced detail and textures.
  • Text: Slightly better, but still unreliable.

Stable Diffusion XL

  • Download : SDXL (huggingface)
  • Training Image Resolution: 1024×1024
  • Ideal KSampler Input Image Resolution: 1024×1024, 1152x896, 896x1152, 1216x832, 832x1216, 1344x768, 768x1344, 1536x640, 640x1536
  • Other Ksampler Settings: Steps 20, CFG 4, Euler, Normal
  • Approx. VRAM Required: ~8–12 GB
  • Realism: Strong cinematic quality, photorealistic textures.
  • People: Handles faces, skin, and anatomy with high fidelity.
  • Landscapes: Delivers rich depth, dramatic lighting, and realism .
  • Text: Improved over earlier SD versions; better legibility with prompt care.

Stable Diffusion 3.5

  • Download : SD3.5 (huggingface)
  • Training Image Resolution: ~1024×1024
  • Ideal KSampler Input Image Resolution: 1024×1024 (dynamic sizes)
  • Other Ksampler Settings: Steps 20, CFG 4, Euler, SGM_Uniform
  • Approx. VRAM Required: ~12–16 GB
  • Realism: Professional; great prompt adherence, diverse lighting.
  • People: More realistic, diverse facial features, fewer biases.
  • Landscapes: High resolution, cinematic but occasionally stylized.
  • Text: Not ideal; better adherence but still error-prone.

FLUX.1 Schnell

  • Training Image Resolution: ~1024×1024
  • Ideal KSampler Input Image Resolution: 1024×1024 (dynamic sizes)
  • Other Ksampler Settings: Steps 4, CFG 1, Euler, Simple
  • Approx. VRAM Required: ~13–33 GB
  • Realism: Very good, faster results, especially with inorganic subjects.
  • People: Weaker than Dev; less refined anatomy.
  • Landscapes: Solid but less detailed than Dev.
  • Text: Struggles. Schnell version doesn’t handle text well.

FLUX.1 Dev

  • Training Image Resolution: ~1024×1024
  • Ideal KSampler Input Image Resolution: 1024×1024 (dynamic sizes)
  • Other Ksampler Settings: Steps 20, CFG 1, Euler, Simple
  • Approx. VRAM Required: ~23–24 GB
  • Realism: Excellent realism, near Midjourney-level photorealism.
  • People: Strong anatomy and hands accuracy.
  • Landscapes: More detailed than Schnell, great for nature and environment.
  • Text: Better than Schnell, but still not perfect. Works for simple/prominent text.

DreamShaper 8

  • Download : Dreamshaper 8 (civitai) | Dreamshaper 8 (huggingface)
  • Training Image Resolution: 512x512 (SD1.5 Base)
  • Ideal KSampler Input Image Resolution: 512x512
  • Other Ksampler Settings: Steps 20, CFG 8, Euler, Normal
  • Approx. VRAM Required: ~4-5 GB
  • Realism: Artistic yet grounded; improved detail over base models.
  • People: Better anatomy and stylization than SD1.5.
  • Landscapes: Flexible; suitable for both creative and typical scenes.
  • Text: Comparable to SD1.5. Garbled unless simple.

AbsoluteReality

  • Download : AbsoluteReality (civitai) | AbsoluteReality (huggingface)
  • Training Image Resolution: 512x512 (SD1.5 Base)
  • Ideal KSampler Input Image Resolution: 512x512
  • Other Ksampler Settings: Steps 20, CFG 8, Euler, Normal
  • Approx. VRAM Required: ~8-12 GB
  • Realism: Highly photorealistic, precise textures and lighting.
  • People: Excellent facial features, diverse, consistent results.
  • Landscapes: Good photorealism landscapes when prompted.
  • Text: Poor; not designed for text generation.

DreamShaper XL

  • Download : Dreamshaper XL (civitai) | Dreamshaper XL (huggingface)
  • Training Image Resolution: 1024x1024 (SDXL Base)
  • Ideal KSampler Input Image Resolution: 1024×1024, 1152x896, 896x1152, 1216x832, 832x1216, 1344x768, 768x1344, 1536x640, 640x1536
  • Other Ksampler Settings: Steps 8, CFG 2, DPM++ SDE, Karras
  • Approx. VRAM Required: ~10-14 GB
  • Realism: Cinematic-level texture/detail; strong across domains
  • People: High accuracy in faces and anatomy, strong depth.
  • Landscapes: Exceptional; cinematic lighting and composition.
  • Text: Better than earlier, but still error-prone. Text legibility not guaranteed.

Image Comparisons

Prompt, "a cat riding a skateboard down the stairs"

a cat riding a skateboard down the stairs

Prompt, "a breathtaking alpine valley at sunrise"

a breathtaking alpine valley at sunrise

Prompt, "a person reading a newspaper"

a person reading a newspaper

Prompt, "a person wearing high tech scifi armor"

a person wearing a high tech scifi armor