Grok 4.5 Launches With Focus on Coding and AI Agents, Intensifying Race With OpenAI and Anthropic
Elon Musk‘s artificial intelligence company has unveiled Grok 4.5, the latest version of its flagship large language model, describing it as a faster, more affordable system designed for coding, software engineering and autonomous AI agents rather than general consumer chat.
The release marks one of the company’s biggest AI announcements since its restructuring under SpaceXAI and comes as competition among leading AI developers accelerates. OpenAI, Anthropic and Google have all introduced increasingly capable models in recent months, with enterprise customers placing greater emphasis on coding performance, reasoning and agentic capabilities.
The company said Grok 4.5 is its “strongest model” to date and has been optimized to handle complex programming tasks and long-running workflows that require AI systems to plan, execute and adapt with limited human intervention.
These so-called AI agents are emerging as one of the industry’s fastest-growing applications, allowing models to write code, analyze data, perform research, and automate business processes.
Unlike previous Grok releases that primarily targeted users of the X social media platform, Grok 4.5 places greater emphasis on enterprise developers through application programming interfaces, or APIs, allowing businesses to integrate the model directly into software products and internal workflows.
The company announced that developers can begin building applications using the new model immediately. Musk described Grok 4.5 as an “Opus-class” model while emphasizing improvements in speed and efficiency compared with rival frontier AI systems.
SpaceXAI also highlighted lower operating costs, pricing the model at approximately $2 per million input tokens and $6 per million output tokens, making it one of the more competitively priced frontier models for enterprise users.
“Grok 4.5 is served at fast-model speeds of 80 TPS. Combined with twice greater token efficiency than the latest leading models at the same tasks, the model delivers intelligent results to you more quickly and at far lower costs,” the company wrote.
The launch reflects a broader shift in the AI industry away from consumer chatbots toward specialized models capable of completing complex work independently. SpaceXAI said Grok 4.5 was built specifically with those workloads in mind, targeting software engineering, technical knowledge work and autonomous task execution.
The release also strengthens Musk’s challenge to OpenAI and Anthropic, whose models continue to dominate enterprise AI adoption. While Grok has gained visibility through its integration into X, the latest version aims to compete directly in the commercial AI market by emphasizing performance per dollar instead of simply raw benchmark scores.
Axios reported that Grok 4.5 trails the very top-performing models in some evaluations but offers an attractive combination of speed, capability and lower inference costs. The company said the model was “trained across tens of thousands of NVIDIA GB300 GPUs, with training and stability techniques designed for large-scale runs.”
Nvidia remains the dominant supplier of hardware powering frontier AI systems, with demand for its processors continuing to surge as developers build increasingly sophisticated models.
