xAI launched Grok 4.7 on September 21, a 2.1-trillion-parameter model that keeps its predecessor's price tag while adding an ingredient no competitor can replicate: internal engineering records from SpaceX.
Grok 4.7 ships with 2.1 trillion parameters, a 500,000-token context window, and a knowledge cutoff of May 2026. It accepts text and image input and produces text output. According to xAI, the model spends more time reasoning through hard problems and checks its own answers more often than Grok 4.6 did, alongside what the company describes as its strongest safety guardrails to date.
The model is already available through Cursor, Grok Build, and the xAI API, and GitHub has begun rolling it out to GitHub Copilot users. xAI also trained Grok 4.7 to natively understand the Grok Bot harness, which the company says makes it behave more naturally in both conversational work and general-knowledge tasks.
The most unusual part of the launch is where some of the training data came from: xAI supplemented the corpus with internal SpaceX engineering records, including Starlink satellite telemetry, manufacturing logs, and failure analyses. It's a type of proprietary data no competing lab has any way to replicate, and it reflects the increasingly tight integration between companies in Elon Musk's ecosystem.
Musk said Grok 4.7's overall performance is roughly comparable to Claude Opus 5.0, though he acknowledged its multimodal capabilities still need work. Trade press coverage was more measured: several outlets described the launch as a bigger model that arrives late to the frontier conversation, given that Anthropic, OpenAI, and Google had all already updated their flagship models in the preceding weeks.
Carlos Montiel is an enterprise AI solutions architect. He implements LLMs, Agents, RAG, and orchestrators for companies across Guatemala and Latin America. Reach out for a consultation.
Contact Carlos Montiel