Home / Directory / AI infrastructure / Bare-metal & clusters / Lambda
Lambda
GPU cloud and on-demand instances aimed at researchers and companies that want direct access to accelerated machines.
Researchers and smaller platform teams that want GPU instances quickly and can manage the software stack.
Enterprises that need reserved multi-thousand-GPU programs, or teams that want a managed training platform.
Verdict
Lambda is a great fit for teams that need a GPU machine or a small cluster and will install the training or serving stack themselves. On-demand pricing is straightforward and often friendlier than a full hyperscaler maze; larger reservations still need a clear SKU conversation. Researchers and builders who want machines this week tend to like the simplicity; it is bare-metal and clusters, not a managed training studio.
Score Breakdown
How Lambda scores in the categories that matter to its buyers.
Pricing
Lambda publishes self-serve On-Demand Cloud rates per GPU-hour. Prices below are USD from Lambda’s public instances pricing as of September 2026. Multi-GPU nodes use the listed per-GPU rate times GPU count. 1-Click Clusters and longer reserved capacity differ from these on-demand rates.
| SKU | Config | Price / GPU / hr | What stands out |
|---|---|---|---|
| NVIDIA B200 SXM6 | 1x | $6.99 | Self-serve on-demand; 8x lists at $6.69 |
| NVIDIA H100 SXM | 1x | $4.29 | 8x SXM lists at $3.99 per GPU-hour |
| NVIDIA H100 PCIe | 1x | $3.29 | Lower than SXM single-GPU rate |
| NVIDIA A100 SXM 80GB | 8x | $2.79 | Per GPU on 8x SXM node |
| NVIDIA A100 40GB | 1x | $1.99 | SXM and PCIe 40GB share this list rate |
| NVIDIA GH200 | 1x | $2.29 | Grace Hopper Superchip |
| NVIDIA A10 | 1x | $1.29 | Smaller inference / fine-tune SKU |
| NVIDIA Quadro RTX 6000 | 1x | $0.69 | Lowest published on-demand GPU |
| 1-Click Cluster H100 (16 GPUs) | 2 weeks-1 year | $6.16 | Reserved cluster list; sales path for larger |
| 1-Click Cluster B200 (16 GPUs) | 2 weeks-1 year | $9.86 | Reserved cluster list; volume tiers lower |
The Field at a Glance
Where Lambda ranks among AI infrastructure vendors we reviewed, by Overall Score and relative typical engagement cost.
Lambda is strongest for researchers and smaller platform teams that want GPU instances quickly and can manage the software stack. It lands mid-pack among the peers we plotted, at 7.3. Relative typical engagement cost is mid-pack versus the peers on this chart. On this broader parent-category chart, Applied Digital leads on Overall Score at 7.6.
Use-case matrix
| Use case | Fit | Notes |
|---|---|---|
| On-demand GPU instances | Strong | Simple and usable. |
| Small research clusters | Mixed | Fine if you run the stack. |
| Hyperscale reserved programs | Poor | Look at a larger GPU cloud. |
| Managed training studio | Poor | You bring the orchestration. |
Who it’s for
Good fit
- Research groups that need machines this week
- Teams comfortable owning the CUDA stack
- Burst training that does not need a multi-year reserve
Poor fit
- Platform orgs buying reserved tens of thousands of GPUs
- Teams that want a job console and nothing else
- Buyers who need deep enterprise contracting on day one
Review Excerpts
Below are excerpts from public reviews. Paid reviews and pay-for-play sites such as Clutch were excluded.
“All of this does require either using the OpenAI APIs, or the Hugging Face APIs, or renting something like a Lambda Labs GPU server and running a 33B model (if you use FastChat, you get an OpenAI compatible API).”
“LambdaLabs GPU cluster provides internode bandwidth of 3.2Tbps: I personally verified it in a cluster of 64 nodes (8xH100 servers) and they claim it holds for up to 5k GPU cluster.”
“I was CPU-limited on the LambdaLabs 1xH100 machines because of this.”
“Lambda Labs advertises 99.9% uptime SLA on reserved instances—and they actually deliver it. In 18 months of production on Lambda, I have seen exactly 2 unplanned outages, both resolved within 40 minutes with automatic incident reports.”
“When my persistent storage corrupted during a critical fine-tuning run last October, I had a senior engineer on a Zoom call within 90 minutes. That experience alone justified the premium.”
“Create account → select GPU → SSH or Jupyter ready in 90 seconds. Pre-installed Docker images for PyTorch, TensorFlow, JAX. Persistent storage volumes (50GB-10TB) attached automatically.”
Company Facts
- Founded
- 2012 (CB Insights)
- Headquarters
- San Jose, California, United States (CB Insights)
- Total funding raised
- $5.574B (CB Insights)
- Key investors
- Mubadala Capital, US Innovative Technology Fund, TWG Global, Credit Agricole, Citigroup (CB Insights)
- Website
- lambda.ai
Key Leadership
Stephen Balaban
Co-founder and CTO
Methodology
This page is an independent evaluation of Lambda for buyers comparing options in bare-metal & clusters. AI Industry Reviews accepts no sponsorships, advertising, or pay-for-placement fees. Lambda did not pay for this review.
What we scored
The headline number is an Overall Score on a 0-10 scale. Eight criteria fall under it in two groups.
- Buyer outcomes (for bare-metal & clusters)
- Instance access
- SKU clarity
- Cluster scale
- Self-managed fit
- Company & commercial
- Innovation &
- product leadership
- Project management &
- communication
- Pricing
- Contract fairness
Pricing measures whether the price looks fair for the value delivered, including packaging and renewal friction that show up in real buying cycles.
Score Composition
| Input | Weight | What it covers |
|---|---|---|
| Reviews | 40% | A proprietary read of what practitioners say about likes, complaints, and day-to-day use, including public review sites, forums, and private chat rooms. Paid reviews and pay-for-play sites such as Clutch are out of scope. |
| Product | 35% | Hands-on look at screens and workflows. |
| Pricing | 15% | Whether the price looks fair for what you get. |
| Docs & training | 10% | Docs, tutorials, and training material. |
How we balanced the evidence
The Overall Score is the simple average of the eight criteria. Recommendation language follows that score and the fit pattern described above.
Scope
Lambda is graded here as bare-metal & clusters. Criteria scores can move as more review volume and product checks are added.