Cost formula
TPM capacity = TPM limit × 60 minutes; Max monthly tokens at limit = daily token cap × 30 (if daily cap binds); Cost at capacity = achievable tokens × blended price per token
Calculator
Rate-limit / concurrency planner
Enter provider RPM/TPM (and optional daily) caps plus your concurrency and latency. CentsPerToken shows the binding constraint, sustainable peak RPM, and the API cost if you run at that capacity.
Capacity
Limits below desired peak
105.3RPM capacity
Binding: TPM limit · headroom -12%
Capacity RPM105.3
Desired peak120
Queue wait est.8.4s
Suggested concurrency4
Req / day @ cap75,789
Cost @ capacity$2,518.11
Limits
Quotas & demand
Edit provider-tier caps and peak demand to find the binding limit.
Gauge
Concurrency capacity
1,900 tokens / request · binding TPM limit
Concurrency load4 / 50
Suggested concurrency to sustain peak: 4. Utilization 8% of your max concurrency cap.
From RPM cap500.0
From TPM cap105.3
From concurrency1500.0
From RPD—
From TPD—
Queue wait estimate8.4s
At capacity on GPT-5 mini: $2,518.11 / month · $0.001108 / request · 2,273,684 requests.
Provider tiers vary by account and change often—treat RPM/TPM fields as editable planning inputs, not live quota. Updated 2026-07-31. Approximate static list prices for planning only. Always verify against each provider’s official pricing page before production budgeting.
Guide
How to get value from this calculator
Identify which provider rate limit binds your workload—RPM, TPM, daily caps, or concurrency—and estimate cost at full capacity utilization. Rate limits determine whether you can spend your budget even when you have one. Plan concurrency and batching before launch to avoid throttling.
Why it matters
A generous budget is useless if TPM caps prevent you from serving traffic. Rate-limit planning connects infrastructure tier upgrades to revenue capacity and maximum possible API spend.
How to use it
- Enter your provider tier limits: requests per minute, tokens per minute, and daily caps.
- Set average input and output tokens per request for your workload.
- Identify whether RPM or TPM binds first at your request pattern.
- Watch the capacity gauge to see which limit binds first.
- Calculate maximum achievable throughput under the binding limit.
- Multiply achievable volume by token pricing to see cost ceiling at limits.
- Compare limits across providers or tiers if you are hitting throttles in production.
Planning tips
- Output tokens count toward TPM on most providers—high output requests hit limits faster.
- Batching small requests reduces RPM pressure but increases latency.
- Multiple API keys or org tiers do not always multiply limits—check provider policy.
- Request a limit increase early if projected traffic exceeds TPM within six months.
- Upgrade the tier that saturates the capacity gauge—not necessarily the one with higher list limits.
- Cost at capacity is an upper bound—actual spend is usually lower unless you run flat out.
Frequently asked questions
Which limit should I plan around?
Whichever binds first at your tokens-per-request and requests-per-minute pattern. This tool helps identify RPM vs TPM vs daily cap binding.
Do rate limits affect pricing?
Limits do not change per-token price but cap how many tokens you can buy per minute. Upgrading tiers often raises both limits and may change effective rates.
Are limit values exact?
Enter limits from your provider dashboard or documentation. Defaults in the tool are planning examples—your tier may differ.
Is cost at capacity a realistic budget?
It is a theoretical maximum if you saturated limits 24/7. Use usage budget planner for realistic traffic-based estimates.