SpaceXAI's Grok 4.7 available on Vercel AI Gateway with four reasoning levels (low/medium/high/xhigh) to tune latency-depth tradeoff; 40% discount through Sept 27.
September 24, 2026
Summary
Unified API endpoint (`spacexai/grok-4.7`) works across AI SDK, OpenAI-compatible Chat Completions, and coding agents (fx, Codex, eve)—reduces integration friction when testing new reasoning models. Configurable reasoning levels let you optimize inference cost and speed per use case.
Why it matters
Unified API endpoint (`spacexai/grok-4.7`) works across AI SDK, OpenAI-compatible Chat Completions, and coding agents (fx, Codex, eve)—reduces integration friction when testing new reasoning models. Configurable reasoning levels let you optimize inference cost and speed per use case.
Implementation verdict
Drop-in replacement for existing OpenAI-compatible calls or AI SDK streams. Requires AI Gateway API key (provisioned via `vercel ai-gateway setup`). Ready now—pricing reflects provider cost with zero platform markup, and supports Zero Data Retention + prompt training disallow. Start with playground or `npx eve@latest init` for quickest agent setup.
Sources
Dev Signal
Get briefs like this in your inbox — free, every weekday.
100+ sources compressed into one 4-minute read. Ranked, cited, implementation-ready.