Illustration for: The Real Cost Of AI Tokens Is Nothing Like The List Price

The Real Cost Of AI Tokens Is Nothing Like The List Price

A new survey of AI token economics finds realized enterprise costs on frontier models run $3-12 per million tokens even as list prices have collapsed to under $1, because agentic workflows multiply token usage 50-500x per task.

By the Numbers

~$0.97/M tokens
List price (blended index)
$3-12/M tokens
Realized enterprise cost
50-500x per task
Agentic usage multiplier
20-25%
Firms with mature AI FinOps
TC
By the AI Desk
Edited by Trace Cohen · Early-stage VC & angel · Founder, New York Venture Partners
1 min read
ShareXLinkedInEmail

THE RUNDOWN

1

The LLM Token Expenditure Index fell to roughly $0.97 per million tokens this month, its lowest level since the index launched -- yet enterprise bills keep rising, the clearest sign yet that list price and actual cost have become disconnected metrics.

2

Agentic workflows -- retries, retrieval, orchestration, observability -- multiply raw token usage by 50 to 500 times per completed task, meaning the headline per-token rate card is closer to a floor than a meaningful predictor of what a company actually pays.

3

Only 20-25% of companies in a May 2026 McKinsey survey had mature AI FinOps capabilities, and managing token cost and usage is now the single top challenge FinOps teams report -- a measurement gap, not just a spending one.

4

Two buyers on the identical vendor rate card can land two to three times apart in cost per completed outcome, because routing choices, prompt size, retry logic and model-tier selection do more to determine total cost than the advertised price ever does.

TC

The VC Read · Trace's Take

Trace Cohen

The number every founder pitching AI margin improvement needs to defend isn't the list price per token, it's the realized cost per completed task -- and the survey's own finding that two buyers on the same rate card land two to three times apart tells you routing and retry logic matter more than whatever discount you negotiated. The diligence item: ask any AI-native startup for cost-per-outcome, not cost-per-token, because only a quarter of companies can even measure that number reliably today.

Analysis

A new survey of enterprise AI spending finds the per-token list price and the actual realized cost of running AI at scale have become almost unrelated numbers. The LLM Token Expenditure Index fell to about $0.97 per million tokens this month, its lowest level since the index launched -- while enterprise AI bills at frontier-tier volume actually land between $3 and $12 per million tokens once real usage patterns are counted.

The gap is structural, not a pricing anomaly: agentic workflows -- retries, retrieval steps, orchestration overhead and observability logging -- multiply raw token consumption by 50 to 500 times relative to a single simple query, according to the survey. That means the advertised rate card functions as a floor on cost, not a reliable predictor of it. Two companies buying from the identical vendor at the identical list price can end up two to three times apart in cost per completed task, driven entirely by how they route requests, how large their prompts run, and how aggressively they retry failed calls.

That means the advertised rate card functions as a floor on cost, not a reliable predictor of it.

The measurement problem compounds the spending problem: only 20% to 25% of companies in McKinsey's May 2026 Enterprise AI FinOps survey reported having mature cost-management capabilities for their AI spend, and managing token cost and usage is now the single most commonly cited top challenge among FinOps teams. For any startup pitching AI-driven margin improvement, or any GP diligencing one, the realistic per-task cost figure -- not the list price the vendor quotes -- is the number that determines whether the unit economics actually work, and most companies still can't measure it with confidence.

ShareXLinkedInEmail

Key Sources

2 sources

Reported by HPCwire · Analysis by Value Add Pulse.

← Back to Pulse

THE WIRE in your inbox— Tech, startup & VC news with Trace's take. Free, no spam.