> ## Documentation Index
> Fetch the complete documentation index at: https://docs.sailresearch.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Provider comparison

> Compare sandbox pricing, capabilities, and machine limits.

| Provider | vCPU | Memory | Max runtime | Memory snapshots | Start/resume time | Maximum machine size |
| - | - | - | - | - | - | - |
| Sailboxes | \$0.015 **active** vCPU·h | \$0.008 **active** GiB·h | No fixed limit | Yes | \<2s | 64 vCPUs, 256 GiB RAM |
| E2B | \$0.0504 reserved vCPU·h | \$0.0162 reserved GiB·h | 24 hours (Pro) | Yes | \<1s | 8 vCPUs, 8 GiB RAM |
| Daytona | \$0.0504 reserved vCPU·h | \$0.0162 reserved GiB·h | No fixed limit | Yes | \<500ms | 4 vCPUs, 8 GiB RAM |
| EC2 On-Demand | \$0.063 reserved vCPU·h† | \$0.007875 reserved GiB·h† | No fixed limit | No | Minutes | 64 vCPUs, 512 GiB RAM |
| Modal | \$0.071 usage/request\* vCPU·h | \$0.024 usage/request\* GiB·h | 24 hours | Yes | \<500ms | 64 vCPUs, 336 GB RAM |
| Cloudflare | \$0.072 **active** vCPU·h | \$0.009 reserved GiB·h | No fixed limit | No | \<3s | 4 vCPUs, 12 GiB RAM |
| AWS AgentCore | \$0.0895 **active** vCPU·h | \$0.00945 **active** GB·h | 8 hours | No | 1s | 2 vCPUs, 8 GB RAM |
| Vercel | \$0.128 **active** vCPU·h | \$0.0212 reserved GB·h | 24 hours (Pro/Enterprise) | No | \<1s | 32 vCPUs, 64 GB RAM |

\*Modal bills the greater of requested and used CPU and memory. Bursting beyond
the request depends on available capacity. These are Sandbox rates; Modal
counts one physical CPU core as two vCPUs.

†The r6i.16xlarge in Oregon costs \$4.032/hour on demand. The CPU and RAM
figures each divide that same host price by its capacity; they are not
separate charges to add together.