Pricing

AI2IN bills metered compute in INR (with GST), for sandboxes running in Mumbai (ap-south-1).

Rate card

Resource	Rate
vCPU	₹4.00 / vCPU-hour
Memory	₹1.00 / GB-hour

You're billed for what a sandbox actually consumes while running. A paused sandbox uses no compute and isn't billed for it. Usage is metered continuously and shown on the Usage page in the dashboard.

Example

A standard code sandbox is 1 vCPU + 0.5 GB RAM. Running it actively for an hour:

1 vCPU-hour × ₹4.00  +  0.5 GB-hour × ₹1.00  =  ₹4.50 / hour

Most agent workloads are bursty — the sandbox is mostly idle between calls — so real spend is a fraction of the always-on figure.

Quotas & caps

Every team has server-enforced limits so costs stay predictable:

• Concurrent sandboxes — how many can run at once.
• Monthly vCPU-hours cap — a ceiling on trailing-30-day compute.
• Budget cap — an optional ₹ spend ceiling.

Hitting a hard cap returns 429. Adjust caps under Limits in the dashboard (team admins), or ask us to raise them.

Managing spend

• Pause long-lived sessions between turns — the biggest lever.
• Watch the Usage page and set a budget cap as a backstop.
• Idle sandboxes are auto-reaped, so abandoned ones don't keep billing.

See the dashboard for current plan details and to add a payment method.