Cost Forecasting

Real-time cost tracking and estimation before you send requests. See exactly what you'll pay with transparent, no-markup pricing.

What is Cost Forecasting?

Cost Forecasting shows you the exact cost of AI requests before and after you make them. No surprises, no markup - just transparent pricing.

How It Works

📸 Cost Display

Cost Tracking in Lab Compare

Lab Compare showing estimated costs before sending and actual costs after responses

📸 SCREENSHOT NEEDED: cost-display.png
What to capture: Lab Compare interface with cost information visible
Show: Estimated cost display before sending, actual cost after response, per-model costs, total cost for the comparison
Tips: Set up a Lab Compare, show costs clearly visible in the UI

Before You Send

  • Estimated Cost - See approximate cost based on prompt length
  • Per-Model Pricing - Compare costs across models
  • Token Estimates - View estimated token usage

After Response

  • Exact Cost - Precise cost of the request
  • Token Breakdown - Input and output tokens
  • Running Total - Session cost tracking

Where You See Costs

Lab Chat

  • Model selector shows cost per 1K tokens
  • Estimated cost before sending
  • Actual cost after response
  • Session total in header

Lab Compare

  • Cost per model before sending
  • Side-by-side cost comparison
  • Total cost for all models
  • Cost-per-quality analysis

Lab Batch Compare

  • Total estimated cost
  • Per-model breakdown
  • Cost summary after completion

Pricing Details

How We Calculate

  • Input tokens - Your prompt
  • Output tokens - AI response
  • Provider rates - Direct from AI companies
  • No markup - We pass through exact costs

Typical Costs

  • GPT-3.5-Turbo: ~$0.001 per request
  • GPT-4-Turbo: ~$0.01-0.03 per request
  • Claude 3.5 Sonnet: ~$0.015-0.025 per request
  • Gemini 1.5 Pro: ~$0.01-0.02 per request

Cost Management Tips

Choose Right Model

  • Use cheaper models for simple tasks
  • Save expensive models for complex work
  • Test with multiple models to find best value

Optimize Prompts

  • Shorter prompts = lower costs
  • Be specific to get shorter responses
  • Avoid unnecessary context

Monitor Usage

  • Check session totals regularly
  • Review team usage in settings
  • Set personal budgets

Team Billing

All costs are:

  • Tracked per user
  • Billed to team account
  • Viewable in team settings
  • Transparent and detailed

Next Steps

Previous: ← Image Generation | Next: Edit Response →

results matching ""

    No results matching ""