xAI Unveils Grok 4 Fast, Offering 98% Cost Reduction for Comparable Performance

xAI has introduced Grok 4 Fast, a new large language model designed for enhanced cost-efficiency and performance, now available to users and developers. The model, which builds on the capabilities of Grok 4, aims to democratize advanced AI by significantly lowering operational costs while maintaining high-level reasoning. This development marks a strategic move by xAI to scale its reinforcement learning efforts, as suggested by industry observers.

According to Haider., a prominent voice on social media, Grok 4 Fast ranks among the top AI models due to several key features. "98% cheaper than Grok 4," he stated in a tweet, highlighting the model's economic advantage. This substantial cost reduction is achieved through optimizations that allow Grok 4 Fast to use approximately 40% fewer "thinking tokens" on average compared to Grok 4 for similar tasks.

Grok 4 Fast boasts multimodal reasoning capabilities, enabling it to process and understand various forms of information. It features a substantial 2 million token context window, allowing for extensive and complex interactions. The model also integrates real-time web and X (formerly Twitter) search, enhancing its ability to provide up-to-date information and insights.

The API pricing for Grok 4 Fast is set at $0.20 per 1 million input tokens and $0.50 per 1 million output tokens for requests under 128k tokens, making it highly competitive. xAI has developed a unified architecture for Grok 4 Fast, allowing it to seamlessly switch between quick responses and deep reasoning based on the query's complexity. This design reduces latency and token costs, making it suitable for a wide range of real-time applications.

Independent evaluations, such as those by Artificial Analysis, have positioned Grok 4 Fast as having a state-of-the-art price-to-intelligence ratio among publicly available models. xAI stated that Grok 4 Fast was trained end-to-end with tool-use reinforcement learning, enabling it to intelligently decide when to leverage external resources like web browsing or code execution. The model is available to all users on grok.com, iOS, and Android apps, with free access offered for a limited time on platforms like OpenRouter and Vercel AI Gateway.