← Providers
Tensorlake
provider: "tensorlake"
Bills active CPU (floor 20% of allocation); this is the full-use price; Pro plan is lower.
Price per hour
| 1 vCPU · 2 GB | $0.100 |
| 2 vCPU · 4 GB | $0.200 |
| 4 vCPU · 8 GB | $0.400 |
| 8 vCPU · 16 GB | max 4 vCPU |
| 16 vCPU · 32 GB | max 4 vCPU |
| 32 vCPU · 128 GB | max 4 vCPU |
Source: www.tensorlake.ai/pricing, checked 2026-11-04.
Adapter
- Supports
- Exec + streaming · Files · Pause · Custom image
- Max size
- 4 vCPU · 16 GB
- Routing
- Chosen when it's the cheapest fit, or as a fallback
setup
npm i @sandboxrouter/sdk @tensorlakeai/tensorlake
export TENSORLAKE_API_KEY=...use Tensorlake
import { Router } from "@sandboxrouter/sdk";
const router = new Router();
const sbx = await router.create({ vcpu: 2, memoryGb: 4 }, { provider: "tensorlake" });
await sbx.exec("uname -a", { onStdout: (c) => process.stdout.write(c) });
await sbx.kill();