What is a good LLM cost-to-revenue ratio?

It varies by model, but AI-native SaaS often targets LLM spend under 10% of revenue, with 5% or less considered healthy. Above 20% usually signals you need cheaper models, caching, or higher pricing.

Why show the ratio against COGS as well as revenue?

Cost as a share of revenue tells you margin pressure; cost as a share of COGS tells you how dominant inference is within your delivery costs. A high COGS share means LLM optimization is your biggest lever.

How does the target slider work?

You pick a target percentage of revenue. The tool computes the LLM spend that ratio implies and the dollar and percentage reduction needed from your current spend to get there.

Does this include fine-tuning or embedding costs?

Include any model-provider spend you want counted in the LLM spend field — inference, embeddings, fine-tuning. The ratio is only as complete as the spend figure you enter.

Is my data sent anywhere?

No. The calculator runs entirely in your browser. Nothing you enter is uploaded, stored or logged.

LLM Cost-to-Revenue Ratio Calculator

LLM cost-to-revenue ratio calculator

One number tells you whether your AI feature is a margin problem: the share of revenue it consumes. Enter your monthly LLM spend, revenue, and total COGS, and this tool returns the cost-to-revenue ratio, the cost-to-COGS share, and — with the target slider — the exact spend reduction needed to hit your goal.

How it works

cost_to_revenue = llm_spend / revenue
cost_to_cogs    = llm_spend / total_COGS
target_spend    = revenue × target_ratio
reduction       = current_spend − target_spend

The revenue ratio is the headline metric investors and finance teams watch. The COGS ratio reveals how much of your delivery cost is inference — if LLM spend is 60% of COGS, model optimization is your single biggest margin lever. The target slider converts a goal into an action: a dollar amount of spend you must remove through cheaper models, caching, shorter outputs, or smarter routing.

Benchmarks and tips

Under 10% of revenue is a common target for AI-native products; 5% or less is comfortable.
Above 20% is a warning sign — revisit model choice, cache aggressive prompts, and cap output length.
Raise price before cutting quality. If the feature drives real value, pricing power often fixes the ratio faster than shaving tokens.