SpaceXAI's Grok 4.6 achieves a score of 61 on the Artificial Analysis Intelligence Index, matching the performance of GPT-5.6 Sol and demonstrating improved agentic capabilities at a lower cost. This version marks a 5-point increase over Grok 4.5 and a 23-point increase compared to Grok 4.3, positioning SpaceXAI alongside OpenAI in the intelligence frontier, just behind Anthropic.

Grok 4.6 enhances capabilities for long-running agents and supports more ambitious interactive and visual tasks. It can manage complex workflows, including researching topics, analyzing information, and developing applications.

Grok 4.5 has launched, described as xAI's smartest model yet and trained alongside Cursor for coding and agentic tasks. A build-off was conducted where Grok 4.5, GPT-5.5, Claude Opus 4.8, and Fable 5 created the same interactive apps, with results measured for latency and cost.

A battle royale simulation with eleven LLMs revealed that Grok 4.1 Fast won 43% of its matches, outperforming competitors. The cheapest model achieved 27 times more cost-effective wins than the most expensive one.

Grok, developed by Elon Musk's xAI, replaced Anthropic's Claude in Project Maven and facilitated over 2,000 munitions strikes against Iran in 96 hours during Operation Epic Fury. The Pentagon confirmed this shift in a legal brief following a contract dispute with Anthropic.

Grok 4.3 offers REST API and gRPC support for developers. Pricing details and documentation are available on the xAI website.

The US government has deployed Elon Musk's AI chatbot Grok as a nutrition bot on the newly launched website RealFood.gov, which promotes protein-centric dietary guidelines. Grok has faced criticism for providing inappropriate advice, including suggestions for rectal use of vegetables.