SpaceXAI shipped Grok 4.7 on the 22nd, positioning it for coding, professional knowledge work and longer-duration tasks. Artificial Analysis gave it an Intelligence Index score of 46, two points above Grok 4.6 and enough to place SpaceXAI fourth among the compared frontier AI labs, behind Claude Fable 5.1, GPT-6 Astra and Claude Opus 5. Grok 4.7 scored 1,657 Elo on AA-Briefcase and 1,695 Elo on GDPval-AA, while its coding-agent score reached 56 on the Coding Agent Index, nine points above Grok 4.6 and ahead of GPT-5.6 Sol. Pricing remains $2 per million input tokens and $6 per million output tokens, with cache hits at $0.50 and a 500,000-token context window. However, Artificial Analysis measured roughly 81,000 output tokens per Intelligence Index task, compared with 36,000 for Grok 4.6 and 27,000 for GPT-6 Astra, increasing the likely cost per completed task. The release retains the broader availability, safety work and Nasdaq-100 developments previously associated with the launch. Elon Musk said the model uses a 2.1-trillion-parameter base and training that included SpaceX company data. He also said Grok 4.8 would finish training within a week and expects Grok 5 to reach artificial general intelligence, while SpaceXAI's division has reported a $1.26 billion loss.