Which is better, Northflank or AWS Lambda MicroVMs?
It depends on how busy the sandbox is and where it must live. For agents that mostly wait, Runtime bills only measured CPU, so the benchmark costs $0.64 there, against $1.11 on Northflank and $4.20 on Lambda. Pick Northflank for CPU-bound work that keeps every core busy, where its low allocated rate beats Runtime, or for GPUs. Pick Lambda MicroVMs for short bursts inside an AWS VPC.
At a glance
| Northflank | AWS Lambda MicroVMs | Runtime | |
|---|---|---|---|
| Isolation | microVMs for CPU workloads, gVisor for GPU workloads | Firecracker microVM | Firecracker microVM, own kernel |
| CPU | $0.01667 per allocated vCPU-hour | $0.0997 per baseline vCPU-hour | $0.025 per vCPU-hour of measured CPU, with a small floor |
| Memory | $0.00833 per GB-hour | $0.0132 per GB-hour | $0.0075 per reserved GiB-hour |
| Lifetime | Runs until you stop it | State kept up to 8 hours | No time limit: runs while it works, or persistent |
| Where | Its regions, or your own AWS, GCP or Azure account | AWS regions | Dedicated servers Runtime operates, one US region |
| GPUs | L4, A100 and H100, billed by the second | No | No |
About six times the CPU rate
The two meters work the same way. Each charges a price per vCPU for every second the sandbox exists, so the gap is purely the rate. Lambda's $0.0997 per vCPU-hour is about six times Northflank's $0.01667. Both can run inside AWS, since Northflank deploys into your own account too. Lambda additionally speaks IAM and your VPC natively.
Cost for the same job
2 vCPU and 4 GiB (4 GB on both rivals), 60 seconds a run, 20 CPU-seconds of work, 1,000 runs:
- CPU
- 1,000 × 60 s / 3,600 × 2 × $0.01667
- $0.556
- Memory
- 1,000 × 60 s / 3,600 × 4 × $0.00833
- $0.555
- CPU
- 1,000 × 60 s × 2 × $0.0000276944
- $3.32
- Memory
- 1,000 × 60 s × 4 × $0.0000036667
- $0.88
- CPU
- 1,000 × 20 s / 3,600 × $0.025
- $0.14
- Memory
- 1,000 × 60 s / 3,600 × 4 × $0.0075
- $0.50
When both CPUs stay busy for the whole minute, the job costs $1.33 on Runtime, $4.20 on Lambda, and $1.11 on Northflank. That is the one case where Northflank's low allocated rate beats measured billing.
Which one to pick
- Northflank for CPU-bound work that keeps every core busy, or GPUs.
- Lambda MicroVMs for short bursts inside an AWS VPC.
- Runtime for agents that mostly wait: the lowest bill of the three, and a sandbox created in one call.
See Runtime vs Northflank and Runtime vs AWS Lambda MicroVMs.
Runtime Cloud
For an agent that waits on a model, Runtime is 42% cheaper than Northflank and 85% cheaper than Lambda MicroVMs. 1,000 one-minute runs of a 2 vCPU, 4 GiB sandbox with 20 CPU-seconds of work each cost $1.11 on Northflank, $4.20 on AWS Lambda MicroVMs, and $0.64 on Runtime.
To start, sign in with npx withruntime login and use the free usage. The
migration guide helps move code from Northflank or Lambda.
Sources
Northflank rates were checked 25 October 2026. Lambda MicroVMs rates were checked 25 October 2026.
Facts on this page were checked on 26 October 2026.