X.ai, Elon Musk's artificial intelligence venture, aims to compete in the generative AI chatbot space alongside OpenAI's ChatGPT and Google's Gemini. As a result, X.ai has introduced an updated version of its AI model, Grok-1.5. This updated model provides more features compared to the old version of Grok. The company claims that Grok-1.5, with improved reasoning abilities, launched with a context length of 128,000 tokens, surpassing its previous model.
Grok-1.5, especially excels in coding and mathematical tasks, demonstrating better reasoning abilities than its predecessor, Grok-1, by more than double on the MATH benchmark. Additionally, it shows an improvement of over ten percent on the HumanEval test, which is better than the previous version in programming language generation and problem-solving skills. It also outperforms various benchmarks such as MMLU, GSM8K, and HumanEval, including Mistral Large AI model. Additionally, xAI plans to introduce several new features in Grok 1.5 in the coming days. Although the model is not available yet, it will be available for testers and existing Grok users on the X platform in the coming days.
It is believed that the integration of Grok's new version will provide X users with a better user experience. Users will be able to solve their queries directly on the platform rather than redirecting to a web browser for answers. Furthermore, the two versions of Grok 1.5 chatbot - fun and regular, make the use of the chatbot even more interesting. The regular mode provides clear answers to questions, while the fun mode explains information in a more humorous way. So far, Grok is only available to X's premium users, meaning users will have to pay to use the AI chatbot services. However, this will bring X a step closer to becoming an everything app.