LLM Token Costs Drop Below $1 Benchmark
The average cost per million LLM tokens has fallen to 97 cents, continuing a significant months-long decline across the industry.
Key takeaways · 2
- 01
Average token costs have dropped below $1 per million.
- 02
Costs have fallen sharply since peaking at $2.07 in May.
Token Costs Decline
According to a report by Alex Harring at CNBC, the LLM Token Expenditure Index shows that the average cost per million tokens has fallen to 97 cents. [1] This current pricing marks a continuation of a sharp, months-long decline. [1] Prior to this drop, the index had hit a high of $2.07 on May 28. [1]
What it means
The reduction in LLM token costs to under a dollar per million highlights a rapid commoditization of inference over the past several months. While the initial spike in May to $2.07 suggested potential pricing pressure, the subsequent drop indicates improved efficiencies or increased competition driving costs down. What the sources don't address: whether this cost reduction applies uniformly across all model tiers or is driven primarily by smaller, more efficient models.
Decreasing token costs significantly lower the barrier for deploying LLM applications at scale. Practitioners can expect improved unit economics for data-intensive AI workloads.
Why it matters
Turn this story into practical AI skill after launch.
Get the release link for daily sessions built around your role and industry.
Join the waitlistHow this developed
1 September 2026
LLM Token Costs Drop Below $1 Benchmark
1 September 2026
Event created from source cluster.