Open-weight and cheap models now carry most production AI traffic, so a premium model's introductory price matters more to buyers than its frontier claims
Price and efficiency, not frontier capability, now decide model adoption. Open-weight and cheap models are taking the volume of production tokens. Premium models are competing on introductory rates and on vendor-run benchmarks that independent measures do not yet confirm. The clearest signal this week is the split between usage and money. Vercel's October 2026 AI Gateway index reported that open-weight models took 71.5% of tokens and 23% of spend. Most of the work flowing through that gateway runs on open-weight models, while most of the money still goes to proprietary ones. Teams are sending routine, high-volume work to whatever is cheapest that does the job and paying premium prices only where they judge the premium is earned. The premium vendors appear to be responding on price rather than capability. Google opened its top model to everyone at introductory rates.
72% Open-weight share of Vercel AI Gateway usage and spend, October 2026
over 50% DeepSeek's claimed token reduction for V4.1 Flash-1017 at unchanged pricing
35% Share of paid Vercel AI Gateway teams using Jev within two weeks
Signal Pro adds the evidence behind the argument, how this week sat against previous weeks, the policy read for every market and what it changed for planning.
Get Signal Pro →The week's read in your inbox every Monday. Unsubscribe anytime.
A research seat runs Mireto's research on any organization in your plan's market and helps you prepare the conversation. Partner firms deliver the result to their clients in their own name.