Elon Musk’s Grok Bot will no longer run only on SpaceX’s own AI models. It will also use models from rivals, including Anthropic’s Claude Opus 5.5, Midjourney and Suno, Musk said in a post on X on Wednesday.

“Going forward, @SpaceX will use the best back end model for any given task, including Claude Opus 5.5, MidJourney, Suno and other leading APIs. Whatever is most likely to give you the best outcome,” Musk wrote.

Grok Bot is the AI agent app from SpaceXAI, the AI business Musk has folded into SpaceX. It launched in August. Each Bot works on its own computer and can carry out several tasks in parallel. Musk did not say which tasks would go to which outside model.

Rivals behind the bot

Musk launched xAI in 2023 to compete with OpenAI and has spent heavily on computing to train its Grok models. Anthropic, Midjourney and Suno build models for text, images and music. The change came a day after users reported problems reaching Grok on its mobile and web services, City AM reported. SpaceXAI gave no official cause.

Anthropic released its cheaper Opus model in September, weeks before its planned stock market listing. This week, Musk said SpaceXAI will take the name SpaceXSI.

A crowded agent market

xAI has kept adding features to Grok Bot. Since late September, teams can publish a shared Bot that colleagues use in the app or in Slack, according to its product changelog.

Grok Bot competes with personal AI agents such as Meta’s Muse, which reportedly passed 3 million weekly users, and Instinct. Some early users have deleted such agents over privacy worries, Business Insider reported on Tuesday. One of them told the publication he still uses Grok Bot, but has not connected it to his email.