What an AI Pilot Actually Costs in 2026
Honest numbers for a first AI project: the assessment, the plan, the build, and the running costs people forget to budget.
5 min read · Mapleforge Labs
The four cost blocks
A realistic first AI project has four blocks: finding the right problem (assessment), specifying the solution (plan), building it (implementation), and running it (subscriptions, tokens, maintenance). Skipping the first two is how businesses end up paying build prices for the wrong tool.
The numbers (our published prices)
A Readiness Assessment at Mapleforge Labs starts at $1,500 — a written report and prioritized roadmap, with 50% credited toward a follow-on build within 60 days. A Strategy & Implementation Plan starts at $2,500. A typical first build runs $3,500–$8,000 quoted per project; integration with an existing legacy system typically runs $5,000–$15,000. Businesses that prefer hourly help pay $150/hour for advisory sessions; individuals pay $95/hour for 1:1 coaching (or $320 for a four-session pack). Team training runs from $1,200 for a half-day workshop.
The running costs people forget
Budget for the subscriptions and token/API usage your pilot consumes, the staff time to verify outputs during the first months, and one review at 90 days to decide — with numbers — whether to scale, fix, or stop. A pilot that can't be evaluated can't be defended. Start with the free 20-minute discovery call; if a pilot doesn't make sense for you yet, we'll say so and tell you what to do instead.
Want this applied to your organization?
Start with the free discovery call — or read about the services these guides come from.