xAI's Grok 4.3 and Build 0.1 models now live on Cloudflare AI Gateway
Unified billing, 1M-token context, and a coding model at $1/M tokens
Elon Musk's xAI has deepened its partnership with Cloudflare, putting the full Grok model lineup — Grok 4.3, Grok Build 0.1, Grok Imagine, Grok Aurora, and Grok Voice — onto Cloudflare's AI Gateway. Announced June 4, 2026, the integration goes beyond a simple listing: developers now access Grok models through Cloudflare's unified control plane for AI API traffic, which handles logging, caching, rate limiting, spend controls, and billing in a single dashboard.
Grok 4.3 stands out as the most cost-effective 1M-context model on the Gateway at $1.25 per million input tokens ($0.20 cached) and $2.50 per million output tokens. Grok Build 0.1, a software engineering model released in beta on May 28, offers always-on reasoning, tool calling, and structured outputs at over 100 tokens per second for $1/M input and $2/M output. Developers still need an xAI API token, but billing is consolidated onto a single Cloudflare invoice. This positions Cloudflare as a foundational 'substrate' for deploying AI agents at scale, aligning with its vision for the agentic web.
- Grok 4.3 offers a 1M-token context window at $1.25/M input and $2.50/M output, the cheapest 1M-context model on Cloudflare AI Gateway
- Grok Build 0.1 is a dedicated coding model with 256K context, 100+ tokens/sec, and always-on reasoning — priced at $1/M input, $2/M output
- Unified billing through Cloudflare eliminates separate invoices for xAI usage, while developers still authenticate with an xAI API token
Why It Matters
Simplifies multi-model AI deployment with consolidated billing and global infrastructure, accelerating developer productivity.