Kimi K3 and Kimi K3 Fast with ZDR and US-based providers now on AI Gateway

By Steven Van · 2026-07-27

Vercel's AI Gateway adds US-based providers and Zero Data Retention for Moonshot AI's Kimi K3 and its faster Kimi K3 Fast variant.

Vercel's AI Gateway from US-based providers, including Baseten and Fireworks, with Zero Data Retention (ZDR) supported for both. Running on US infrastructure lets teams with data residency and compliance requirements meet those constraints while still using the model.

The gateway routes across multiple providers under a single moonshotai/kimi-k3 model ID, handling failover and provider selection automatically.
Kimi K3 Fast trades a higher per-token cost, about 50% more than the base model, for lower latency. Request it with the speed option on the base model, or call moonshotai/kimi-k3-fast directly.
Setting inferenceRegion restricts inference to US data centers, at roughly a 10% premium over standard pricing.
ZDR can be turned on for every request from the AI Gateway dashboard, or per request with zeroDataRetention.

Read the original announcement

Vercel

The platform for frontend developers — deploy, preview, and scale web apps and AI agents with zero config.

View Vercel →

Read the original announcement →