AI Run Planner

Model

Prompt tokens / request

Output tokens / request

Requests / minute (avg)

Peak factor ×

Active hours / day

Days / month

Cache hit %

Retry %

Base latency P50 (ms)

Latency per 1k tokens (ms)

Throughput per instance (RPS)

Concurrency cap / instance

Currency

Cost / 1k requests

—

Est. monthly cost

—

Latency P50 / P95

—

Needed instances

—

Prices are placeholders. Update per model as needed.

Model comparison

Model	Input $/1k tok	Output $/1k tok	Cost/1k req	Monthly cost	P50/P95 ms	Instances