Estimate your monthly API cost
Estimate the tokens behind API requests, chat sessions, or agent runs, then apply your model’s rates.
How the estimate worksThree choices shape every estimate
Workload type controls how calls and history are counted. Agent presets are examples; replace their values with your application’s measurements.
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
A user’s task can involve several model calls. Count what each call receives and produces, including the calls made by parallel workers.
Include context sent again
For an assumed five-call run, a 1,000-token prefix sent on every call contributes 5,000 input tokens before new messages or tool results.
Count each model call
Planning, worker tasks, summaries, and retries may add calls to the workflow.
Check the largest context
Monthly totals estimate the bill; the largest individual context is compared with the model limit. Passing that check does not validate the whole workload.
Use rates for your model
Replace the example rates and context limit with the values that apply to your selected model.
Compare API rates using the same call assumptions. For Product Plan, bring peak traffic and request sizes when discussing capacity.
Continue exploring
1,000 calls a day · 30 days · 50% cached input
What will GPT‑OSS 120B cost?
Compare four rate cards, including ours, using your request volume, token usage, and cache assumptions.
57.40public index, 8B SH
- 3B SH
- 51.53
- Retrieval & Classification
- 66.4
- Tools & Automation
- 84.0
Decision Index 0.3 public index, self-scored
Standard One SH scores every option in one pass
Open-weight decision models that return a probability for every option in one forward pass: 51.5 and 57.4 on the Decision Index 0.3 public index, self-scored with the official kit, with the detail and the limits.
TPMa rate with provider-specific rules
- Capacity
- PTU · MU · GSU
- Billing
- Units · GPU-hours
- Limits
- Windows · concurrency
Definitions before price comparisons
LLM capacity and plan terms defined
Definitions for inference capacity units, billing terms, workload estimates, and subscription limits, with provider-specific meanings kept explicit.