Accenture and Amazon Web Services Propose a Three-Phase Sequence for Optimizing AI Token Costs, with Four Clients Reporting Reductions of 40% to 85%
After enterprises scale their deployment of artificial intelligence (AI), the price per token decreases but total expenditure increases. This article outlines six token cost optimization levers jointly proposed by Accenture and Amazon Web Services (AWS), ordered into three phases: configuration changes, architectural changes, and infrastructure investment. It reports actual results from four clients: an insurance client and a mining client using only a single caching layer saw reductions of approximately 40%, while a large enterprise and a telecommunications operator applying multiple levers combined achieved reductions of 85% and 68%, respectively. The core argument is that the optimization goal is not to minimize tokens at all costs, but to match the appropriate level of intelligence to each business outcome.
A counterintuitive phenomenon often arises after enterprises scale their deployment of AI: while the price per token is falling, total token expenditure is rising. This occurs because cheaper tokens stimulate greater adoption, and agentic workflows split a single user interaction into multiple model invocations. A practice article jointly published by Amazon Web Services (AWS) and Accenture points out that in response to this situation, the correct optimization goal is not to reduce the number of tokens at any cost, but to match the appropriate level of intelligence to each business outcome—finding the corresponding point between cost, quality, and risk.
The article categorizes six optimization levers into three phases based on implementation complexity. Phase 1 involves pure configuration changes without altering infrastructure: prompt caching reuses repetitive system instructions and tool definitions, with design goals aimed at reducing cached input token costs by up to 90%; batch inference asynchronously processes non-urgent tasks in batches, costing 50% less than on-demand calls. Phase 2 involves architectural changes: intelligent prompt routing directs requests to different-sized models within the same model family, reportedly reducing costs by up to 30% without sacrificing accuracy; Retrieval-Augmented Generation (RAG) optimization reduces irrelevant context entering the model through hierarchical chunking, metadata filtering, and reranking. Phase 3 requires infrastructure and training investment: semantic caching stores semantically similar previously answered requests in vector storage; model distillation transfers knowledge from large models to small models, reportedly increasing speed by up to 500% and reducing costs by up to 75%.