Context economics
LLM Conversation Context Cost Calculator
Visualize how repeated chat history, tool definitions, and long conversations compound token spend.
Your scenario
Adjust the assumptions. Results update instantly.
Monthly volume of multi-turn sessions.
Instructions included at the start of every model call.
Function schemas and tool instructions repeated with each call.
Average new user message and retrieved context.
Average generated response retained in history.
Number of user-assistant exchanges.
How frequently older history is replaced by a compact summary.
Tokens retained after each history compaction.
Share of calls receiving discounted system and tool tokens.
Methodology
How this calculator works
Rebuild the full input context for every conversation turn.
Apply prompt-cache pricing to stable system and tool content.
Compare full history with periodic summarization to expose avoidable context spend.
What makes this useful
This models billed context turn by turn and compares full history with scheduled summarization.
Frequently asked questions
Why does conversation cost grow every turn?
Each model call usually includes prior messages, so old tokens are billed repeatedly while new context continues accumulating.
Does a larger context window reduce cost?
A larger window prevents truncation but does not make tokens free. Retaining more history can increase spend even when the model accepts it.
When should conversation history be summarized?
Summarize when the cost and latency of resending raw history exceed the quality value of retaining every message verbatim.
Continue your analysis
Caching Break-Even Calculator
Find the reuse rate and traffic volume required for caching to save money.
Open toolFallback Chain Calculator
Measure reliability cost across primary calls, retries, fallbacks, and failed tasks.
Open toolUsage Limit Calculator
Set a usage allowance that is generous to customers and safe for gross margin.
Open toolLLM Conversation Context Cost Calculator uses current model prices and your operating assumptions to estimate business impact. Treat the output as a planning model, then replace defaults with p50, p95, and p99 telemetry from your own product.
Related tools: All AI SaaS tools · LLM pricing database · Full margin simulator