The AI Taxonomy.
Terms  /  Measurement  /  Recall Cost Ratio
09 · Measurement

Recall Cost Ratio

The cost of retrieving stored context vs. the cost of re-generating it from scratch.

Recall Cost Ratio determines whether your memory system is actually saving money. Storing and retrieving context has a cost: storage infrastructure, retrieval computation, and the tokens used to inject recalled context into the prompt. If that cost exceeds the cost of simply re-establishing the context from scratch (by re-explaining to the model), the memory system is a net negative. A good Recall Cost Ratio means retrieval is significantly cheaper than regeneration — usually by an order of magnitude or more.

Example
A team's memory system stores project context and retrieves it at the start of each session. Cost of retrieval: ~500 tokens for the retrieval query + 2,000 tokens of recalled context = 2,500 tokens. Cost of re-onboarding without memory: ~6,000 tokens of re-explanation to reach the same level of project understanding. Recall Cost Ratio: 2,500/6,000 = 0.42. The memory system is 58% cheaper than starting from scratch. If the ratio were above 1.0, the memory system would be more expensive than re-onboarding — a sign to optimize or remove it.

More in Measurement

Context Efficiency

Useful retained state per token consumed.

Token ROI

The measurable value produced per token consumed. The business metric for Context Efficiency.

Context Utilization Rate

The percentage of input context that the model actually uses for its output. Low rate = waste.

Waste Ratio

The proportion of tokens consumed that produce no useful output. The metric behind Token Bleed and Token Burn.

Iteration Velocity

The speed at which a human-AI pair can move from idea to tested output.

Recovery Cost

The time and tokens required to get a session back on track after a failure — State Loss, Execution Hallucination, or Plan Drift.

Quality-Per-Token

A measure of output quality relative to tokens consumed. Distinct from Token ROI (value) — this measures correctness, completeness, and relevance per unit of cost.