AI Cloud Pricing
Clear, straightforward pricing for On-Demand Instances, 1-Click Clusters, Private Cloud, and serverless Inference.
Contact us for reserved capacity at our lowest prices.
1-Click Clusters Pricing
On demand. Self-serve. Short or Long-term. As low as $1.85 for H100 or $2.99 for B200 for committed usage. Learn more
GPU
| 16-1.5k NVIDIA HGX B200 GPUs | ||
| Commitment | As low as (per GPU-hour) | |
| On-demand | 1 week+ | $3.79 |
| Reserved | 1 year | $3.49—Contact us |
| Reserved | 2 years | $3.29—Contact us |
| Reserved | 3 years | $2.99—Contact us |
| NVIDIA H100 GPUs | ||
| Commitment | As low as (per GPU-hour) | |
| On Demand | 1 week-3 months | $2.69—contact us |
| Reserved | 3-6 months | $2.29—contact us |
| Reserved | 6-12 months | $2.19—contact us |
| Reserved | 1 year | $1.85—contact us |
On-Demand Cloud Pricing
Access high-power GPUs when you need them. Pay by the minute with no egress fees. Learn more
Choose your GPU configuration:
| GPUs | VRAM/GPU | vCPUs | RAM | STORAGE | PRICE/GPU/HR* |
| On-demand 8x NVIDIA H100 SXM | 80 GB | 208 | 1800 GiB | 22 TiB SSD | $2.99 |
| Reserved 8x NVIDIA H100 SXM (min. 32 GPUs) | 80 GB | 208 | 1800 GiB | 22 TiB SSD | CONTACT SALES |
| On-demand 8x NVIDIA A100 SXM | 80 GB | 240 | 1800 GiB | 19.5 TiB SSD | $1.79 |
| On-demand 8x NVIDIA A100 SXM | 40 GB | 124 | 1800 GiB | 5.8 TiB SSD | $1.29 |
| On-demand 8x NVIDIA Tesla V100 | 16 GB | 88 | 448 GiB | 5.8 TiB SSD | $0.55 |
*plus applicable sales tax
| GPUs | VRAM/GPU | vCPUs | RAM | STORAGE | PRICE/GPU/HR* |
| On-demand 4x NVIDIA H100 SXM | 80 GB | 104 | 900 GiB | 11 TiB SSD | $3.09 |
| On-demand 4x NVIDIA A100 PCIe | 40 GB | 120 | 900 GiB | 1 TiB SSD | $1.29 |
| On-demand 4x NVIDIA A6000 | 48 GB | 56 | 400 GiB | 1 TiB SSD | $0.80 |
*plus applicable sales tax
| GPUs | VRAM/GPU | vCPUs | RAM | STORAGE | PRICE/GPU/HR* |
| On-demand 2x NVIDIA H100 SXM | 80 GB | 52 | 450 GiB | 5.5 TiB SSD | $3.19 |
| On-demand 2x NVIDIA A100 PCIe | 40 GB | 60 | 450 GiB | 1 TiB SSD | $1.29 |
| On-demand 2x NVIDIA A6000 | 48 GB | 28 | 200 GiB | 1 TiB SSD | $0.80 |
*plus applicable sales tax
| GPUs | VRAM/GPU | vCPUs | RAM | STORAGE | PRICE/GPU/HR* |
| On-demand 1x NVIDIA GH200 | 96 GB | 64 | 432 GiB | 4 TiB SSD | $1.49 |
| On-demand 1x NVIDIA H100 SXM | 80 GB | 26 | 225 GiB | 2.75 TiB SSD | $3.29 |
| On-demand 1x NVIDIA H100 PCIe | 80 GB | 26 | 225 GiB | 1 TiB SSD | $2.49 |
| On-demand 1x NVIDIA A100 SXM | 40 GB | 30 | 220 GiB | 512 GiB SSD | $1.29 |
| On-demand 1x NVIDIA A100 PCIe | 40 GB | 30 | 225 GiB | 512 GiB SSD | $1.29 |
| On-demand 1x NVIDIA A10 | 24 GB | 30 | 226 GiB | 1.3 TiB SSD | $0.75 |
| On-demand 1x NVIDIA A6000 | 48 GB | 14 | 100 GiB | 512 GiB SSD | $0.80 |
| On-demand 1x NVIDIA Quadro RTX 6000 | 24 GB | 14 | 46 GiB | 512 GiB SSD | $0.50 |
*plus applicable sales tax
Transparent pricing
For On-Demand instances, pay by the minute only for what you use. Get discounted pricing for long-term commitments.
NVIDIA HGX B200 available as low as $2.99/GPU/hour.
Flexible commitments
We know your AI needs can change—and fast.
Build a flexible reservation tailored precisely to your timeline and budget, without getting locked in to a particular GPU.
Unmatched AI expertise
We're engineers, not resellers. We focus on AI/ML infrastructure with a no-BS approach.
Think high-performance compute, transparent pricing, and support from people who get it.
Private Cloud Pricing
Single tenant, caged clusters for large AI deployments. One-, two-, or three-year contracts for 1,000-64k GPUs and 3.2Tb/s networking.
| INSTANCE TYPE | GPU | GPU MEMORY | vCPUs | STORAGE | NETWORK BANDWIDTH |
| NVIDIA HGX B200 | B200 SXM | 180 GB | 224 | 60 TB local per 8x B200 | 3200 Gbps per 8x B200 |
| NVIDIA HGX H200 | H200 SXM | 141 GB | 224 | 30 TB local per 8x H200 | 3200 Gbps per 8x H200 |
| NVIDIA H100 | H100 SXM | 80 GB | 224 | 30 TB local per 8x H100 | 3200 Gbps per 8x H100 |
As low as $2.99 for B200 with a multi-year commitment. Learn more
Inference API Pricing
Use the latest open-source models, scale effortlessly, and only pay for the tokens you use with no rate limits. Learn more or read the docs
Model type
| Model | Quantization | Context | Price per 1M input tokens |
Price per 1M output tokens
|
| Deepseek-R1-0528 | FP8 | 164k | $0.50 | $2.18 |
| DeepSeek-V3-0324 | FP8 | 164k | $0.34 | $0.88 |
| Qwen-3-32B | FP8 | 41k | $0.10 | $0.30 |
| Llama-4-maverick-17b-128e-instruct-fp8 | FP8 | 1M | $0.18 | $0.60 |
| Llama-4-scout-17b-16e-instruct | FP8 | 1M | $0.08 | $0.30 |
| Llama-3.1-8B-instruct |
BF16 | 131K | $0.025 | $0.04 |
| Llama-3.1-70B-instruct |
FP8 | 131K | $0.12 | $0.30 |
| Llama-3.1-405B-instruct |
FP8 | 131K | $0.80 | $0.80 |
* plus applicable sales tax
| Model | Quantization | Context | Price per 1M input tokens |
Price per 1M output tokens
|
| DeepSeek-llama3.3-70b | FP8 | 131k | $0.20 | $0.60 |
| Llama-3.3-70B-instruct | FP8 | 131K | $0.12 | $0.30 |
| Llama-3.2-3B-instruct |
FP8 | 131K | $0.015 | $0.025 |
| Hermes-3-Llama-3.1-8B |
BF16 | 131K | $0.025 | $0.04 |
| Hermes-3-Llama-3.1-70B (FP8) |
FP8 | 131K | $0.12 | $0.30 |
| Hermes-3-Llama-3.1-405B (FP8) |
FP8 | 131K | $0.80 | $0.80 |
| LFM-40b |
BF16 | 66K | $0.15 | $0.15 |
| Llama3.1-nemotron-70b-instruct | FP8 | 131K | $0.12 | $0.30 |
| Qwen2.5-Coder-32B |
BF16 | 33K | $0.07 | $0.16 |
* plus applicable sales tax
Frequently Asked Questions
What makes Lambda different from other providers?
Lambda is the only cloud provider focused solely on AI. We offer high-performance GPU cloud compute with transparent pricing, and you can get started with 1-Click Clusters with no commitment and no sales call.
What support do you offer?
We offer direct access to leading AI infrastructure engineers who understand your models and requirements—no tiered support queues or generic help desks.
Do you offer Managed Kubernetes and Managed Slurm?
Yes. See our Orchestration page for more details.
What makes Lambda's Inference API different?
Our Inference API features the lowest rates on leading open-source models with no rate limits.
What are 1-Click Clusters?
Lambda's 1-Click Clusters feature 16-512 NVIDIA GPUs, from 1 week to 3 years, joined with Infiniband connectivity.
Is there a minimum commitment for your On-Demand Cloud?
No. Spin up instances when you need them, spin them down when you don’t.
How does Lambda handle security?
Security here isn't just a department; it's a shared mission. We build a culture where everyone understands their part in protecting our systems and—more importantly—our customers' data.
Does Lambda support Enterprises?
Yes. Lambda is an end-to-end and secure AI compute platform: Private Cloud, on-premise GPU servers and desktop workstations, all bolstered by enterprise-grade security (SOC2 Type II compliant), and best-in-class AI-expert support. Learn more