Introducing Grok 4.5
xAI's advanced model for coding, agentic tasks, and knowledge work
Grok 4.5 is xAI’s latest model optimized for coding, agentic tasks, and knowledge work. Trained on extensive datasets covering coding, science, engineering, and math, it demonstrates strong reasoning and outperforms many comparable models in real engineering benchmarks. The training involved tens of thousands of NVIDIA GPUs with advanced data curation and reinforcement learning focused on multi-step software engineering tasks. Grok 4.5 is capable of building complex applications from minimal prompts, including simulations and end-to-end app development. It operates at fast speeds of 80 transactions per second and achieves roughly 4.2 times greater token efficiency compared to similar models, reducing computational costs. Beyond coding, Grok 4.5 is integrated into Grok Build, where it supports complex Excel modeling, PowerPoint diagram creation, and Word document drafting. The model is competitively priced at $2 per million input tokens and $6 per million output tokens. Grok 4.5 is accessible through the xAI console, API, Grok Build, and Cursor platforms, with free usage available for a limited time to encourage adoption.
Sources
- xAIx.ai