# Cost Estimate — {{pipeline_name}}

Generated: {{date}}
Monthly volume: {{monthly_volume}} calls

---

## Per-Call Cost

| Step | Model | Input tokens | Output tokens | Cost/call |
|------|-------|-------------|--------------|-----------|
| {{step_name}} | {{model}} | {{input_tokens}} | {{output_tokens}} | ${{cost_per_call}} |
| **Total** | | | | **${{total_cost_per_call}}** |

## Monthly Projections

| Scenario | Monthly cost |
|----------|-------------|
| No optimization | ${{monthly_no_opt}} |
| With prompt caching | ${{monthly_with_cache}} |
| After all optimizations | ${{monthly_optimized}} |
| Potential savings | {{savings_pct}}% |

## Model Selection Rationale

**{{model}}** selected for this pipeline because:
- {{model_rationale}}

Haiku feasibility: {{haiku_feasibility}}

## Caching Opportunities

{{caching_analysis}}

## Optimization Recommendations

{{optimization_recommendations}}

---

*Generated by `aiwg nlp estimate-cost {{pipeline_name}}` — update after prompt changes*