Vercel published its AI Gateway Production Index for September 2026 (data through August). Open-weight models processed a majority of gateway tokens for the first time, rising from 7% in December to 56% in August, while average token price fell another 23.2%. A special report says GPT-6 Astra took 7.7% of all gateway spend in its first twelve days, roughly twice Fable 5.1 at 3.7%.

Key Takeaways

  • ✓Open-weight models ran 56% of AI Gateway tokens for the first time, up from under 10% in December 2025.
  • ✓Average token price fell another 23.2% in August, the third straight monthly drop; median spend for high-volume teams fell 7.6%.
  • ✓GPT-6 Astra took 7.7% of all gateway spend in its first twelve days, about twice Fable 5.1, and a third of OpenAI spend within two days.
⚡

Key Decision Metrics at a Glance

TL;DR Verdict
🏆 Benchmark RankRank
Open-weight models ran 56% of AI Gateway tokens for the first time, up from under 10% in December 2025.
Full Leaderboard→
💰 Cost & PlansCompare
Average token price fell another 23.2% in August, the third straight monthly drop; median spend for high-volume teams fell 7.6%.
Compare Plans↗
🛠️ Ecosystem & IDERules
Vercel · AI Gateway · Production Index
Setup IDE Rules↗
🧭

Turn your technical choice into a development budget

Compare 40 dev plans & simulate token costs vs $20/mo subscriptions

⚔️ Compare Model→💰 Compare plans & limits↗
Action HubReady to adopt this in production?

Benchmark side-by-side against alternatives, or calculate monthly token cost vs subscription break-even.

⚔️ Launch VS Compare→💳 View Plans↗
Up NextVercel: TypeSafe Jev is the fastest-adopted model in AI Gateway history
Read Now→