In the latest tech updates, xAI has officially launched its lightweight flagship model, Grok4Fast. This model not only achieves significant breakthroughs in performance but also brings surprising reductions in operational costs, demonstrating strong competitiveness.
The core highlight of Grok4Fast is a 40% reduction in computational requirements. This means it can be more efficient when handling complex tasks, especially suitable for applications that require high timeliness. Moreover, the cost reduction per task is as high as an astonishing 98%. For enterprise users, this is undoubtedly a great opportunity to save on expenses, significantly lowering usage costs while maintaining effectiveness.
This model supports a context window of up to 2 million tokens, allowing it to handle larger-scale data with ease. In core data tests, Grok4Fast achieved a high score of 85.7% on the GPQA Diamond benchmark and an excellent score of 92.0% in the AIME2025 professional evaluation. These results demonstrate Grok4Fast's strong performance in the industry, making it capable of handling various complex analytical tasks.
In terms of commercialization, xAI has introduced a tiered pricing plan, with usage costs ranging from $0.05 to $1.00 per million tokens. This offers flexible options for more enterprises. Notably, interested users can experience the powerful features of Grok4Fast for free through platforms such as OpenRouter and Vercel, personally experiencing the charm of this model.
The release of Grok4Fast marks another major advancement for xAI in the field of artificial intelligence. It not only provides users with more efficient solutions but also brings new opportunities for enterprises in cost control. As technology continues to evolve, we look forward to Grok4Fast performing outstandingly in more application scenarios.
