Estimate your monthly API cost.

Choose a request, chat, or agent workload. Adjust the starting values to match your product. This estimates API usage, not subscription capacity.

Get API access How the estimate works

This tool estimates token-billed API workloads. It does not calculate Coding Plan subscription prices or Product Plan allocations for your product.

API requests Chat sessions Agent workflows

Agent workflow structure

Choose a starting workflow, then adjust its model calls, tool results, and context assumptions. A compact loop that alternates model actions and tool observations.

Workflow preset

Runs per monthRuns Instructions and tool definitionsTokens Task prompt tokensTokens Initial context tokensTokens Model rounds per runRounds Tool calls per roundTools Action tokens per round/ round Tokens per tool result/ tool Final output tokensTokens History sent again% Maximum working-history tokens0 = no cap

History sent again is the share of earlier content included in later model calls. It does not change how Standard Thinking stores your data. A history cap of 0 means no cap in this estimate.

Model and cost assumptions

Model context limitTokens Repeat or retry multiplier× Input price per 1M tokens$/1M Output price per 1M tokens$/1M

The multiplier repeats the entire scenario, including model and tool calls. Add model-only evaluation calls to the workflow assumptions instead. Rates are editable assumptions; view API rates before using the estimate for budgeting.

Estimated monthly tokens processed

—Tokens / month

Input tokens: —
Output tokens: —
Model calls: —
Tool calls: —
Largest request context: —
Repeated input prefix: —

Use of context limit—

The estimated request context fits within the selected limit.

Estimated monthly API model cost—

Where the estimated tokens come from

Simple tool loop · per run

Tool results are shown separately. Only the portion sent to the model counts as input tokens.

Breakdown of the current token estimate.
Component	Input tokens per run	Output tokens per run	Tokens produced by tools

Repeated input is counted each time it is sent to the model. This estimate applies the full input rate without a cache discount; it does not predict cache hits.

This is a planning estimate, not metered usage. It does not infer tool compute, storage, network charges, hidden reasoning, or multimodal processing from token counts.

Three choices shape every estimate.

Choose a workload, then replace the starting assumptions with values from your application. Agent presets describe example workflows, not product benchmarks.

What does one workload look like?

Requests counted separately · Chat history · Tools and model calls

API · Chat · Agent

Workflow preset

A compact loop that alternates model actions and tool observations.

Simple tool loop

Model and cost assumptions

The workload multiplier scales the complete scenario, including model and tool calls, to represent retries or repeated runs.

Context limit · rates

Count what the model processes on each call.

Include instructions, earlier messages, plans, and tool results each time they are sent to the model. Count parallel workers as separate calls.

Include context sent again.

A run with five model calls counts its repeated prefix five times, so input grows with the number of calls.

Count each model call.

Planning, worker tasks, summaries, and retries may add calls to the workflow.

Check the largest context.

Monthly tokens help estimate cost. The largest request context helps assess model fit.

Use rates for your model.

Replace the example rates and context limit with the values that apply to your selected model.

Use the estimate to compare API rates. For a Product Plan, share your peak traffic and request-size assumptions with us.

View API pricing Discuss a Product Plan

Groq

$15.75

Fireworks

$13.95

Together

$18.00

Standard Thinking

$4.77

1,000 calls a day · 30 days · 50% cached input

Tool · September 30, 2026

What will GPT-OSS 120B cost?

Compare four rate cards, including ours, using your request volume, token usage, and cache assumptions.

27terms defined

Capacity units: 11
Plan mechanics: 8
Coding windows: 5

A fourth group, reading a rate, adds three more

Guide · September 29, 2026

Capacity and plan terms, defined

A working glossary for capacity plans and coding plans: 27 terms, each quoted from the vendor's own page or from ours, with the ones we have not published marked.

6units for one allowance

Formula public: 1 of 6
No figure: 3 of 6
Open signup: 5 of 6

Six vendors' published units, checked September 29, 2026

Comparison · September 29, 2026

How coding plans count usage: credits, requests, dollars, and windows

A coding plan's price is comparable; its allowance is not. What one unit counts at each vendor, which window refills it, and what the pages leave unsettled.

View all resources