Methodology
Hourly input and output tokens are scaled by active hours, developers, and workdays, then priced independently at the selected model's catalog rates.
Token and API Costs
Estimate monthly API spend for repository agents from active hours, token burn, workdays, and model rates.
Interactive Tool
| Result | Value | How to read it |
|---|---|---|
| Agent hours / month | 440 | Active agent time across the team. |
| Input tokens / month | 220M | Repository, prompt, and tool input volume. |
| Output tokens / month | 35.2M | Generated output volume. |
| Estimated API spend | $3,960.00 | Input and output priced separately. |
| Cost / developer | $792.00 | Monthly API spend divided across developers. |
The model rates are the lowest published catalog prices. Verify the endpoint actually used by the coding product.
Catalog model: Anthropic Claude Fable 5 →
Inputs
| Input | How it is used |
|---|---|
| Active hours | Agent-running hours per developer per day. |
| Team and workdays | Developers and working days included in the month. |
| Hourly token burn | Observed input and output tokens per active agent hour. |
| Model | Supplies sourced token prices. |
Hourly input and output tokens are scaled by active hours, developers, and workdays, then priced independently at the selected model's catalog rates.
Measure token burn from actual agent sessions if possible. Repository indexing, retries, compaction, and tool traces can make coding workloads much heavier than chat.
Boundaries
Catalog values retain their source and freshness on the linked model, provider, benchmark, or comparison page. Editable scenario assumptions are not Model Markets measurements.
Continue the Analysis
| Tool | Next question |
|---|---|
| API vs Subscription | Find the usage level where seat pricing and metered API pricing cross for a team. |
| Monthly AI Spend | Build a bottom-up monthly AI budget from product or team behavior. |
| Prompt Cost | Make the cost of a system prompt, tool schema, or repeated instruction visible before scale. |
Questions
Convert an agent's observed hourly token burn into a monthly cost scenario. It returns daily and monthly input, output, and total API spend for the selected model.
Where the calculation needs model facts, it uses the current Model Markets catalog snapshot updated 2026-09-02. User-entered assumptions remain clearly editable, and unsupported values stay unknown rather than being inferred.
Subscription allowances and overage rules are excluded. Cache discounts require a separate caching scenario. The lowest model price may not match the endpoint used by the agent product. Open the linked canonical records and primary sources before making a production or purchasing decision.