Open-weight models cross 50% of AI Gateway token volume
General
Vercel's AI Gateway routes tens of trillions of tokens monthly between production applications and AI labs, and the company publishes what that traffic shows about enterprise AI usage in its monthly Production Index. The September 2026 report, which draws on gateway data collected through August 2026, includes prior reports from June, July, and August.
What the September index shows
- Open-weight models carried the majority of gateway tokens for the first time, rising from 7% in December to 56% in August.
- The average token now costs less than half its price five months earlier. Price per token dropped 23.2% in August, the third consecutive monthly decline, and the median team paid 7.6% less.
- Fable 5, described as Anthropic's most capable model, lost two-thirds of its share of gateway spend in a single month. Opus 5, priced at half as much, tripled its share. Anthropic still captured 64% of all spend.
- Gemini 3 Flash shed 95% of its share of gateway tokens since May, and more than three-quarte
The report is credited to Amelia Charles and Eric Dodds.
