Most SMBs Are Still Paying Last Year's AI Prices

An audit of 40 small and mid-sized business AI stacks found 68% were paying more than twice current market rates, with a median $1,100 in monthly savings left on the table. Layer3Labs says falling model prices and old contracts are creating a widening gap between what AI should cost today and what many companies still pay.

Why it matters: - AI model prices have dropped 70% to 90% this year, but many SMBs have not adjusted their contracts or usage. - The pricing gap is leaving a median of $1,100 per month in potential savings unclaimed across the audited businesses. - Layer3Labs says the issue is now a budgeting problem, not just a technology problem, because AI has started to behave like a commodity.

What happened: - Layer3Labs reviewed 40 small and mid-sized business AI stacks between February 2026 and September 2026. - 68% of those businesses were paying more than twice the current market rate. - OpenAI cut its flagship pricing by 80% this month. - Most businesses have not gone back to vendors to renegotiate after the price drops.

The details: - Layer3Labs identified three common ways businesses are overpaying. - Many companies route routine work to frontier models even though much of that work could run on cheaper options. - Some “AI-powered” SaaS add-ons resell model usage at roughly four times the underlying cost. - Some annual contracts signed before price declines keep auto-renewing at the old rate because they lack repricing clauses. - Layer3Labs compared the audited stacks with published vendor rates as of September 2026. - The full breakdown is available in the AI overpayment checker. - Layer3Labs has reworked client AI stacks around current pricing and says model costs fell by up to 60% year over year for those clients. - Clients cited in the release include The Metabolic Journal, The Bot Scout and HOA Guide. - Adam Newhouse, CEO of HOA Guide, said a two-day review cut the company’s AI costs by more than half.

Between the lines: - The release argues that many businesses are paying outdated prices because AI costs moved faster than procurement cycles and contract terms. - Vendors generally do not alert customers when their own costs fall, which can leave savings sitting in place until someone audits usage. - The audit suggests the biggest opportunity is not new AI adoption, but re-routing existing work and renegotiating existing agreements.

What's next: - Layer3Labs is offering AI cost and workflow audits to help businesses reset pricing and usage. - The company also continues publishing independent AI model pricing data on its website. - Businesses with older AI contracts may see the fastest savings by checking routing, seat-based markups and renewal clauses.