Nearly two weeks after Elon Musk’s xAI startup opened up the AI model behind Grok to the public, its AI chatbot is set to get an upgrade.
The company announced Grok-1.5 on Thursday and claimed that its latest model can understand longer documents, handle more complex prompts, and perform more advanced reasoning.
While Grok-1.5 appears to be a step up from the original 1.0 with improvements in coding and math skills, its announcement post shows that it still lags behind Google’s Gemini Pro 1.5 AI, OpenAI’s GPT-4, and Anthropic’s Claude 3 Opus in some benchmark tests, while outperforming OpenAI on one key HumanEval test.
Related: Meet Grok: Elon Musk Unveils ‘Spicy’ AI Chatbot Riddled With ‘Sarcasm’ and ‘Humor’
Grok-1.5 scored higher than GPT-4 on the HumanEval benchmark, which consists of 164 challenging programming problems not included in the AI model’s training data. GPT-4 had a score of 67% and Gemini Pro 1.5 scored 71.9%, while Grok-1.5 received 74.1%.
With a score of 81.3% on the MMLU test, which covers knowledge of 57 subjects from an elementary to an advanced level, Grok-1.5 performed close to Google Gemini’s score (83.7%).
It also scored close to GPT-4’s score of 52.9% with a score of 50.6% on the MATH test, a benchmark that covers grade school to high school math competition problems.
Musk stated in a Friday social media post that Grok 1.5 should be available on X, formerly Twitter, by next week.
The X owner has high expectations for the next generation of Grok, writing that the next step after Grok-1.5 will outperform the AI currently available “on all metrics.” Grok 2 is “in training now,” he wrote in the post.
Grok AI is currently only available to those with a $16 a month or higher Premium+ subscription on X.
Musk sued OpenAI, a competitor of xAI, earlier this month and asked for a court ruling that would force OpenAI to make the research and technology behind its AI public.