Cost

fal.ai charges $0.0400/image for Flux Kontext Pro at 1024x1024. Full cost breakdown at scale, billing mechanics, and comparison to competing APIs. Prices verifi

fal.ai charges $0.0600/image for Ideogram V3 at 1024x1024. Full cost breakdown at scale, billing mechanics, and comparison to competing APIs. Prices verified Ju

Replicate charges $0.0900/image for Ideogram V3 at 1024x1024. Full cost breakdown at scale, billing mechanics, and comparison to competing APIs. Prices verified

Lambda charges $1.09/hr for an RTX A6000. Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices verified June 2026.

fal.ai charges $0.0400/image for Recraft V3 at 1024x1024. Full cost breakdown at scale, billing mechanics, and comparison to competing APIs. Prices verified Jun

Replicate charges $0.0400/image for Recraft V3 at 1024x1024. Full cost breakdown at scale, billing mechanics, and comparison to competing APIs. Prices verified

RunPod charges $4.39/hr for an H200 SXM. Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices verified June 2026.

RunPod charges $0.860/hr for an L40S. Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices verified June 2026.

RunPod charges $0.460/hr for an RTX 3090. Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices verified June 2026.

RunPod charges $0.990/hr for an RTX 5090. Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices verified June 2026.

RunPod charges $0.490/hr for an RTX A6000. Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices verified June 2026.

Salad charges $0.0900/hr for an RTX 3090. Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices verified June 2026.

TensorDock charges $0.200/hr for an RTX 3090. Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices verified June 2026.

TensorDock charges $0.450/hr for an RTX A6000. Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices verified June 2026.

Vast.ai charges $0.290/hr for an A40 (range $0.290-$0.600). Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices verified

Vast.ai charges $2.52/hr for an H100 NVL (range $1.87-$3.33). Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices verifi

Vast.ai charges $3.88/hr for an H200 SXM (range $1.97-$7.37). Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices verifi

Vast.ai charges $0.470/hr for an L40S (range $0.430-$1.00). Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices verified

Vast.ai charges $0.240/hr for an RTX 3090 (range $0.0700-$13.33). Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices ve

Vast.ai charges $0.330/hr for an RTX 4080 (range $0.130-$0.800). Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices ver

Vast.ai charges $0.870/hr for an RTX 5090 (range $0.210-$53.33). Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices ver

Vast.ai charges $0.390/hr for an RTX A6000 (range $0.270-$0.670). Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices ve

AWS charges $5.12/hr per A100 80GB (p4d.24xlarge (8x A100)). Spot savings, egress costs, and comparison to GPU-specialist providers. Prices verified May 2026.

AWS charges $6.76/hr per H100 80GB (p5.48xlarge (8x H100)). Spot savings, egress costs, and comparison to GPU-specialist providers. Prices verified May 2026.

Azure charges $6.98/hr per H100 80GB (Standard_NC40ads_H100_v5). Spot savings, egress costs, and comparison to GPU-specialist providers. Prices verified May 202

CoreWeave charges $2.70/hr for an A100 80GB. Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices verified May 2026.

CoreWeave charges $6.16/hr for an H100 80GB. Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices verified May 2026.

Crusoe charges $1.80/hr for an A100 80GB (range $1.65-$1.95). Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices verifi

Crusoe charges $3.90/hr for an H100 80GB. Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices verified May 2026.

fal.ai charges $0.0250/image for Flux Dev at 1024x1024 at 28 inference steps. Full cost breakdown at scale, billing mechanics, and comparison to competing APIs.

Replicate charges $0.0250/image for Flux Dev at 1024x1024 at 28 inference steps. Full cost breakdown at scale, billing mechanics, and comparison to competing AP

Together AI charges $0.0154/image for Flux Dev at 1024x1024 at 28 inference steps. Full cost breakdown at scale, billing mechanics, and comparison to competing

Black Forest Labs charges $0.0400/image for Flux Pro 1.1 at 1024x1024. Full cost breakdown at scale, billing mechanics, and comparison to competing APIs. Prices

fal.ai charges $0.0450/image for Flux Pro at 1024x1024. Full cost breakdown at scale, billing mechanics, and comparison to competing APIs. Prices verified May 2

Replicate charges $0.0400/image for Flux Pro at 1024x1024. Full cost breakdown at scale, billing mechanics, and comparison to competing APIs. Prices verified Ma

fal.ai charges $0.0030/image for Flux Schnell at 1024x1024 at 4 inference steps. Full cost breakdown at scale, billing mechanics, and comparison to competing AP

Novita AI charges $0.0010/image for Flux Schnell at 1024x1024 at 4 inference steps. Full cost breakdown at scale, billing mechanics, and comparison to competing

Replicate charges $0.0030/image for Flux Schnell at 1024x1024 at 4 inference steps. Full cost breakdown at scale, billing mechanics, and comparison to competing

Runware charges $0.00060/image for Flux Schnell at 1024x1024 at 4 inference steps. Full cost breakdown at scale, billing mechanics, and comparison to competing

Together AI charges $0.0027/image for Flux Schnell at 1024x1024 at 4 inference steps. Full cost breakdown at scale, billing mechanics, and comparison to competi

Segmind charges $0.0080/image for Flux / SDXL at 1024x1024. Full cost breakdown at scale, billing mechanics, and comparison to competing APIs. Prices verified M

GCP charges $4.50/hr per A100 80GB (a2-highgpu-1g). Spot savings, egress costs, and comparison to GPU-specialist providers. Prices verified May 2026.

GCP charges $9.80/hr per H100 80GB (a3-highgpu-1g). Spot savings, egress costs, and comparison to GPU-specialist providers. Prices verified May 2026.

Lambda charges $2.79/hr for an A100 80GB. Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices verified May 2026.

Lambda charges $3.99/hr for an H100 80GB. Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices verified May 2026.

Modal charges $2.10/hr for an A100 80GB. Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices verified May 2026.

Modal charges $3.95/hr for an H100 80GB. Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices verified May 2026.

Oracle Cloud (OCI) charges $1.50/hr per A100 80GB (BM.GPU.A100-v2.8). Spot savings, egress costs, and comparison to GPU-specialist providers. Prices verified Ma

Oracle Cloud (OCI) charges $2.00/hr per H100 80GB (BM.GPU.H100.8). Spot savings, egress costs, and comparison to GPU-specialist providers. Prices verified May 2

RunPod charges $1.19/hr for an A100 80GB. Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices verified May 2026.

RunPod charges $2.49/hr for an H100 80GB. Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices verified May 2026.

RunPod charges $0.340/hr for an RTX 4090. Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices verified May 2026.

Salad charges $0.990/hr for an H100 80GB. Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices verified May 2026.

Salad charges $0.200/hr for an RTX 4090. Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices verified May 2026.

Runware charges $0.0026/image for SDXL at 1024x1024 at 30 inference steps. Full cost breakdown at scale, billing mechanics, and comparison to competing APIs. Pr

TensorDock charges $0.750/hr for an A100 80GB. Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices verified May 2026.

TensorDock charges $2.25/hr for an H100 80GB. Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices verified May 2026.

TensorDock charges $0.360/hr for an RTX 4090 (range $0.350-$0.370). Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices

Thunder Compute charges $0.780/hr for an A100 80GB. Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices verified May 202

Thunder Compute charges $1.38/hr for an H100 80GB. Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices verified May 2026

Vast.ai charges $0.901/hr for an A100 80GB (range $0.901-$0.901). Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices ve

Vast.ai charges $1.38/hr for an H100 80GB (range $0.900-$1.87). Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices veri

Vast.ai charges $0.670/hr for an RTX 4090 (range $0.363-$1.20). Full cost breakdown, estimated cost per AI image, and comparison to other providers. Prices veri

Compare Flux Dev vs Flux Pro vs Schnell. Real latency benchmarks and cost per image across cloud providers. Choose the right model for your budget.

Real numbers on self-hosting Stable Diffusion: hardware, electricity, DevOps, and Kubernetes vs managed GPU APIs. Break-even analysis, verified May 2026.

Tier 1 inference APIs, Tier 2 GPU rental, and Tier 3 hyperscalers compared by real cost-per-image and cost-per-hour. Prices verified May 2026.

Flux Schnell from $0.0027 to $0.003 per image. Real prices for Flux Schnell, Dev, and Pro across Together.ai, Replicate, fal.ai, Novita. Verified May 2026.

Flux Schnell costs $2.70 per 1,000 images via API. Self-hosted on Salad costs $0.27. Real numbers for every path: inference APIs vs GPU rental. May 2026.