AI BriefingKimi K3 and Kimi K3 Fast Now Support ZDR and Feature US-Based Provider AI Gateway
Key point
Moonshot AI's Kimi K3 models are now supported through US-based providers on Vercel AI Gateway, including ZDR capability.
Details
Moonshot AI's Kimi K3 and the faster-responding Kimi K3 Fast are now available on AI Gateway through US-based providers such as Baseten and Fireworks.
With this update, teams with data residency and compliance requirements can operate the models within US infrastructure. AI Gateway automatically routes across multiple providers to support failover in case of outages, ensuring higher availability and throughput than a single provider.
Key features include:
- Zero Data Retention (ZDR): Supports ZDR, which does not store request data, for data security.
- Kimi K3 Fast: Costs about 50% more than the base model but offers lower latency.
- US Infrastructure Inference: Using the
inferenceRegionsetting, requests can be routed exclusively to US data centers, which incurs an additional cost of about 10%.
Developers can call the models via the AI SDK, or use the Vercel CLI to easily complete coding agents configuration.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
vLLM v0.28.0 Released with Major Optimizations for Kimi-K3 and DeepSeek V4
vLLM v0.28.0 Release: Kimi-K3 and DeepSeek V4 Optimizations, Significant Inference Performance Improvements
Underdog Releases 'Husky', a Model-Specific Inference Engine Up to 4.5x Faster than MLX