Gen AI models tech info (image-gen only)

• Boogu-Image Turbo 10B

Tensors 942

context_refiner 2
double_stream_layers 8
image_index_embedding [5, 3,360]
noise_refiner 2
norm_out 4
ref_image_patch_embedder 2
ref_image_refiner 2
single_stream_layers 32
time_caption_embed 7
x_embedder 2

• Krea2 Turbo 12B 2,048 x 2,048

blocks 28 
first 2
last 4
tmlp 2
tproj 1
txtfusion 49
txtmlp 3

• HiDream O1 Image 7B 2,048 x 2,048

Tensors 1,476

model 1,476

• HiDream 17B

Tensors 1,615

caption_projection 49
double_stream_blocks 16
final_layer 4
p_embedder 4
single_stream_blocks 32
t_embedder 4
x_embedder 2

• ERNIE‑Image 8B

Tensors 409

adaLN_modulation 1
final_linear 2
final_norm 2
layers 36
text_proj.weight [4,096, 3,072]
time_embedding 4
x_embedder 2

• Microsoft Lens 3.8B 1440x1440

Tensors 1,264

img_in 2
norm_out 2
proj_out 2
time_text_embed 4
transformer_blocks 48
txt_in 2
txt_norm 4

• Qwen-Image-2512 19B

Tensors 1,933

model 1,933

• Flux.2 Klein 9B

Tensors 201

double_blocks 8
double_stream_modulation_img.lin.weight [24,576, 4,096]
double_stream_modulation_txt.lin.weight [24,576, 4,096]
final_layer 2
img_in.weight [4,096, 128]
single_blocks 24
single_stream_modulation.lin.weight [12,288, 4,096]
time_in 2
txt_in.weight [4,096, 12,288]

• Flux.2 30B

Tensors 299

double_blocks 8
double_stream_modulation_img.lin.weight [36,864, 6,144]
double_stream_modulation_txt.lin.weight [36,864, 6,144]
final_layer 2
guidance_in 2
img_in.weight [6,144, 128]
single_blocks 48
single_stream_modulation.lin.weight [18,432, 6,144]
time_in 2
txt_in.weight [6,144, 15,360]

• Flux.1 12B

Tensors 780

double_blocks 19
final_layer 4
guidance_in 4
img_in 2
single_blocks 38
time_in 4
txt_in 2
vector_in 4

• Z Image Turbo 9B

Tensors 453

cap_embedder 2
cap_pad_token [1, 3,840]
context_refiner 2
final_layer 4
layers 30
noise_refiner 2
t_embedder 4
x_embedder 2
x_pad_token [1, 3,840]

• SD XL 3B

Tensors 2,515

conditioner 587
first_stage_model 248
model 1,680