Pricing & Commercial
Lambda's pricing is positioned cleanly below hyperscalers and slightly above pure marketplaces. The pricing reflects a dedicated enterprise-tier offering at competitive rates rather than aggressive discounting.
On-demand pricing
Indicative on-demand pricing (varies by region and availability):
| GPU | Lambda on-demand | Hyperscaler list | Vast on-demand |
|---|---|---|---|
| A100 80GB | $1.30-1.80/hr | $4-5/hr | $1.00-1.40/hr |
| H100 80GB | $2.50-3.20/hr | $8-12/hr | $2.00-3.00/hr |
| H200 | $3.30-4/hr | $10-14/hr | $2.50-3.50/hr |
| B200 | $5.50-6.50/hr (limited) | Bespoke | Sparse |
Lambda's on-demand is 60-70% below hyperscaler list and modestly above Vast on-demand. The positioning matches the value proposition — better reliability and operational quality than Vast at a price premium; significantly cheaper than hyperscalers.
Reserved pricing
Reserved discounts:
- Monthly reservation: 10-20% off on-demand.
- Six-month reservation: 25-35% off.
- Annual: 35-45% off.
- Multi-year strategic: bespoke pricing further below annual rates.
Reserved Lambda Cloud H100 capacity can come in at $1.50-2/hour effective, which is genuinely competitive with the larger enterprise neoclouds and significantly below hyperscalers.
Cluster pricing
1-Click Cluster pricing depends on cluster size, duration, and configuration. Indicative:
- 16x H100 cluster: $40-50/hour total ($2.50-3.10 per GPU).
- Larger clusters: per-GPU pricing similar; total scales linearly.
- Reserved cluster discounts available.
The cluster pricing is competitive with CoreWeave / Crusoe for similar cluster sizes; deal terms differ.
vs Competitors
- vs Vast: Lambda costs more but delivers dedicated reliability; not directly competing.
- vs RunPod Secure Cloud: Comparable; small differences in specific GPU pricing.
- vs CoreWeave: Lambda is cheaper for smaller customers but doesn't have the bespoke-pricing advantage at very large reservation scale.
- vs hyperscalers: 50-70% cheaper, depending on commitments.
Enterprise contracts
Lambda offers standard enterprise contract structures:
- Master service agreements.
- Custom pricing on volume.
- SOC 2 compliance posture.
- Dedicated account management for larger customers.
- Direct technical support during deployments.
The enterprise sales motion is mature enough for Lambda to compete for seven- and eight-figure annual deals.
Where Lambda's pricing is competitive
Lambda is most price-competitive for:
- Mid-market customers needing reliable GPU compute without negotiating with CoreWeave at scale.
- Customers wanting on-demand flexibility (Lambda's on-demand product is strong).
- Customers needing 1-Click Cluster training without multi-year commitments.
- Customers transitioning from on-prem Lambda hardware to cloud — natural Lambda-to-Lambda relationship.
Takeaway
Lambda's pricing model is conservative-and-competitive — not the cheapest, not the most expensive, well-positioned for its target enterprise mid-market. The next chapter looks at the infrastructure that supports the offering.