Press release: Stride Launches AI Token Cost Optimization as Worldwide AI Spending Climbs Toward $2.6 Trillion With Little Measurable Return
Engineering teams can now see cost-per-outcome visibility in two weeks and cut 40 to 60 percent of AI costs on routed surfaces, with no measurable quality loss
Stride, the AI-native engineering firm, today announced AI Token Cost Optimization, a key missing piece of the AI ecosystem, that connects enterprise AI spend to the business outcomes it produces. The offering makes its national debut at Ai4 2026, August 4 to 6 at The Venetian in Las Vegas, where Stride is exhibiting at booth 1220.
The launch lands at a moment when the economics of enterprise AI are under scrutiny. Gartner forecasts worldwide AI spending will reach nearly $2.6 trillion in 2026, up 47 percent from 2025. Yet in PwC's 2026 Global CEO Survey of more than 4,400 chief executives, 56 percent reported that AI delivered neither increased revenue nor decreased costs over the past year. Only 12 percent reported both higher revenue and lower costs. The spend is compounding. The clarity is not.
"Every CTO I talk to is getting the same two questions from their board: what is AI costing us, and what is it returning," said Debbie Madden, Founder and Chairwoman of Stride. "Almost nobody can answer both. This is not an overspending problem. It is a visibility problem, and it is solvable."
Stride's AI Token Cost Optimization treats AI unit economics as an engineering discipline. By baselining per-feature token attribution we establish cost-per-outcome metrics, cost per inquiry resolved, per document analyzed, and per transaction processed, within two weeks. Optimization sprints then execute a prioritized backlog: model routing and tiering that sends routine tasks to affordable models and complexity to premium ones, prompt and context restructuring measured by cost per outcome, and eval infrastructure that validates every cost-saving change before it ships. The engagement closes with a governance handoff, so the client's team owns the dashboards, alerting, and review gates after Stride leaves.
The results are specific. Model routing typically delivers 40 to 60 percent cost reduction on routed surfaces with no measurable quality loss. In one recent engagement, Stride cut 36 percent from a $1 million annual AI spend.
"Everyone is debating which model to use. Almost nobody asks when to use which one," said Francisco Martin, CEO of Stride. "The model is the easy part. Routing, evals, and cost-per-outcome instrumentation are what turn AI spend into a line item you can defend. We build that alongside the client's engineers, so it keeps working long after we leave."
The approach is proven where the bar is highest. For one telehealth client, routing 85 to 87 percent of SMS-based patient inquiries to high-speed models produced $360,000 in annual savings.
AI Token Cost Optimization is available now. Engagements begin with a two-to-three-week AI Token Cost Assessment that delivers a prioritized action plan with quantified savings opportunities. Ai4 attendees can meet the Stride team at booth 1220. Learn more at stride.build/ai-token-cost-optimization.
About Stride
Stride is an AI-native engineering firm constructing the agentic enterprise. Since 2014, Stride has served 175+ clients across 350 engagements, with deep expertise in regulated industries such as healthcare and financial services. Stride identifies ROI before designing anything and delivers through fixed-scope engagements. Learn more at stride.build.
Sources: Gartner, "Gartner Forecasts Worldwide AI Spending to Grow 47% in 2026," May 19, 2026. PwC, "29th Annual Global CEO Survey," January 20, 2026.
