Skip to content
1 minute readNewAgent-ready

Price suggestions for products with limited quote history

The frontier model combines a product's quote history with patterns learned from simulated markets to suggest a price and estimated acceptance.

The Frontier price model card on Learned pricing in dark mode, with a suggested price 8.5% above today and the acceptance the model expects.

The frontier model can suggest a price even when a product has little quote history. Its estimates depend on simulated training assumptions and the evidence available for that product.

  • Give a product ID and, optionally, a unit cost. The suggestion uses the product's recent quotes and returns estimated acceptance at the suggested and current prices.
  • The bundled model was trained on simulated markets. Its benchmark results measure performance on simulations, not a guarantee of demand or profit in your market.
  • Suggestions stay within bounded ranges around observed or current prices. Review the estimate before using it; the model does not change quoted prices.
  • Use the Frontier model card on Learned pricing, GET /api/frontier-model/suggest, or the suggest_frontier_price MCP tool.