Every turn resends the chat β see why a long conversation costs far more than its words.
Resending history every turn means long chats cost far more than the words in them. Cut the repeated prefix with caching β see the break-even on the prompt cache calculator β and check the running total fits the window on the context window calculator.
The Cost Per Conversation Calculator estimates what a multi-turn chat actually costs to run, not just the price of the new text a user types. You enter the turns per conversation, the new tokens per turn, your conversations per month, and a blended price per million tokens. Because most chat systems resend the growing conversation history on every turn, each turn pays for all prior turns again. The calculator sums that repeated context across turns to produce a per-conversation cost and a monthly total, showing why a chat can cost far more than the tokens it visibly adds.
The key trade-off to watch is conversation length. Since resent context accumulates, cost rises faster than the turn count alone suggests, so long threads dominate your bill. A practical step is to cap history, summarize older turns, or start fresh sessions to keep the resent context small.
Because the model is stateless: every turn you resend the entire prior conversation as input. By turn eight you are paying for the first turn eight times over. Total billed tokens grow roughly with the square of the turn count, so length, not just message size, drives the cost.
Trim or summarise old turns instead of resending everything, use prompt caching so the repeated prefix is billed at a discount, cap conversation length, and route simple turns to a cheaper model. These attack the re-sent context that makes long chats expensive.