Skip to content
2 minute readNewAgent-ready

Price exploration lets learned pricing keep learning

Learned prices now vary slightly from quote to quote, by up to 10% by default, so Last Price can learn what your buyers will pay. Never below cost, always inside your guardrails, and you can turn it off.

Price exploration on the Learned pricing page, switched on with a variation of up to 3% and the explanation that prices never go below cost and stay inside the guardrails

Learned pricing can only measure how buyers react to price when prices move. With price exploration, prices for the same product vary slightly from quote to quote, by up to the percentage you set, so Last Price can learn what buyers will pay.

  • On with learned pricing: it follows your learned pricing switch, and you can turn it off or set the variation (0.5% to 10%) on the Learned pricing page, or with the get_price_exploration and set_price_exploration MCP tools.
  • Steered once there is data: once Last Price has learned enough about a product, the variation moves toward the price it expects to earn most, by up to 10% and never more than 10% past the prices already tried.
  • Limits: never below your unit cost, always inside your own limits and your price guardrails, and only learned and optimizer prices vary. A fixed price or a price from your own functions never does.
  • Per call, not per buyer: each quote draws afresh, and nothing about the buyer is used.
  • Opt one call out: send no_explore in the request options to get the price exactly as computed.
  • Visible: the response and its trace show how much a price moved and from what.