Comfy-org

Comfy-Org/z_image_turbo

Comfy-Org/z_image_turbo

Parameters

6.2B

Context

Downloads

5.5M

Likes

814

Architecture

View on HuggingFace

GPU Requirements

VRAM / RAM / disk estimates by quantization and usage scenario

QuantizationMinimumRecommendedProduction
FP1616bit
19.5 GB
RAM 32.0 · Disk 56.9
35.9 GB
RAM 53.8 · Disk 82.9
133.4 GB
RAM 200.1 · Disk 121.9
FP88bit
12.2 GB
RAM 32.0 · Disk 47.9
28.3 GB
RAM 42.4 · Disk 73.9
125.5 GB
RAM 188.3 · Disk 112.9
INT88bit
12.2 GB
RAM 32.0 · Disk 47.9
28.3 GB
RAM 42.4 · Disk 73.9
125.5 GB
RAM 188.3 · Disk 112.9
AWQ4bit
8.7 GB
RAM 32.0 · Disk 43.6
24.6 GB
RAM 37.0 · Disk 69.6
121.7 GB
RAM 182.6 · Disk 108.6
GPTQ4bit
8.8 GB
RAM 32.0 · Disk 43.7
24.7 GB
RAM 37.0 · Disk 69.7
121.8 GB
RAM 182.6 · Disk 108.7
GGUF4bit
8.9 GB
RAM 32.0 · Disk 43.8
24.8 GB
RAM 37.2 · Disk 69.8
121.9 GB
RAM 182.8 · Disk 108.8

GPU Recommendations

Best ValueBEST

1× RTX 5090

32 GB VRAM

$246/mo

≈ 603.4 tok/s

Effective production throughput ≈ 422.4 tok/s

score 0.6924

Deploy on vast ↗
Cheapest

2× RTX 4070S Ti

32 GB VRAM

$99.0/mo

≈ 0 tok/s (N/A — benchmark unavailable)

Effective production throughput ≈ 0 tok/s

score 0.6402

Deploy on vast ↗
Performance

8× H200 SXM

1128 GB VRAM

$20,440/mo

≈ 5171.7 tok/s

Effective production throughput ≈ 3620.2 tok/s

score 0.2197

Min Complexity

1× RTX A6000

48 GB VRAM

$215/mo

≈ 258.6 tok/s

Effective production throughput ≈ 181 tok/s

score 0.6278

Deploy on vast ↗

💬 Community — real deployment experience

0 Articles · 0 Benchmarks

View Community →

GPUs that fit (single card)

RTX A6000

48 GB VRAM

$215

/mo · vast

RTX 5090

32 GB VRAM

$246

/mo · vast

L40S

48 GB VRAM

$341

/mo · vast

RTX PRO 6000 WS

48 GB VRAM

$482

/mo · vast

RTX PRO 6000 S

48 GB VRAM

$634

/mo · vast

AmciHub

AI Model Deployment & Compute Intelligence — analyze AI model requirements, GPU performance, cloud pricing and deployment costs to find the right deployment solution.

© 2026 AmciHub. All rights reserved. Model → Requirement → GPU → Benchmark → Cloud → Cost → Recommendation