SpaceXAI Launches Grok 4.6 With Focus on Long-Running AI Agents and Coding

SpaceXAI has launched Grok 4.6, its latest AI model, with a focus on long-running AI agents, software development and complex interactive tasks. The company says the model is designed to work through multi-step assignments, including research, data analysis, coding and building applications.

Grok 4.6 builds on Grok 4.5 with a longer supplemental training run and additional training data focused on reasoning, engineering and technical concepts. SpaceXAI said it also used model-generated data, improved its training recipe and expanded reinforcement learning across coding, knowledge work and specialised environments.

"Grok 4.6 achieves frontier intelligence across several agentic coding and knowledge work benchmarks. It matches GPT-5.6 Sol on the Artificial Analysis Intelligence Index, which is a composite score of nine benchmarks," SpaceXAI said in a blog post.

On benchmarks, Grok 4.6 scored 61 on the Artificial Analysis Intelligence Index, matching GPT-5.6 Sol and trailing Fable 5’s score of 62. It scored 69.9% on CursorBench v3.2, 65.9% on DeepSWE v1.1 and 61.3% on FrontierCode v1.1. However, the model did not lead every test, highlighting the increasingly competitive nature of the frontier AI market.

"Grok 4.7 will exceed all current models. That said, Anthropic is a great company and will probably release improved models soon. However, the SpaceX training corpus is so awesome & unique that I would be shocked if any model is better at real-world engineering than 4.7," Elon Musk, SpaceXAI CEO, said.

Musk adds that Grok 4.7 is significantly better than 4.6 and should be ready in 3 to 4 weeks. For now, the company is positioning the Grok 4.6 as a stronger option for tasks that require sustained execution rather than simple question-and-answer interactions.

According to SpaceXAI, Grok 4.6 can research unfamiliar topics, structure applications, implement core functionality and refine projects through multiple rounds of feedback.

The model also showed improvements in visual and interactive work, with xAI saying it can produce stronger first versions of applications and other digital projects.

Grok 4.6 is available through Grok Build, Cursor and xAI’s API, as well as platforms including OpenRouter, Vercel and Cloudflare. API pricing starts at $2 per million input tokens and $6 per million output tokens.

SpaceXAI is also offering twice the included usage on Grok Build and Cursor during the first week of availability, as it pushes the new model toward developers building increasingly autonomous AI applications.

Recently, SpaceXAI released Imagine Image 2.0, a new image generation and editing model designed to produce visuals for professional creative work rather than just one-off AI experiments.