xAI has released an early preview of Grok-2, marking a significant advancement from the previous Grok-1.5 model. Grok-2 introduces frontier capabilities in chat, coding, and reasoning. Alongside it, xAI is also introducing Grok-2 mini, a smaller but equally capable version. An early iteration of Grok-2 has already been tested on the LMSYS leaderboard under the name “sus-column-r,” where it currently outperforms Claude 3.5 Sonnet and GPT-4-Turbo.
Both Grok-2 and Grok-2 mini are now in beta on ๐, with plans to make them available through xAIโs enterprise API later this month.
Benchmarks
The Grok-2 models have been evaluated across a series of academic benchmarks, including reasoning, reading comprehension, math, science, and coding. Both models show significant improvements over the previous Grok-1.5 model, achieving competitive performance levels against other leading models in areas such as graduate-level science knowledge (GPQA), general knowledge (MMLU, MMLU-Pro), and math competition problems (MATH).
Additionally, Grok-2 excels in vision-based tasks, delivering state-of-the-art performance in visual math reasoning (MathVista) and document-based question answering (DocVQA).
Experience Grok with Real-Time Information on ๐
In recent months, xAI has been continuously refining Grok on the ๐ platform. The latest evolution of the Grok experience features a redesigned interface and new functionalities. ๐ Premium and Premium+ users now have access to two new models: Grok-2 and Grok-2 mini. Grok-2 serves as a state-of-the-art AI assistant with advanced capabilities in both text and vision understanding, integrating real-time information from ๐, accessible through the Grok tab in the ๐ app.
Grok-2 mini, though smaller, strikes a balance between speed and answer quality. Compared to its predecessor, Grok-2 is more intuitive, steerable, and versatile across a wide range of tasks, from answering questions to collaborating on writing or solving coding challenges.
In collaboration with Black Forest Labs, xAI is experimenting with their FLUX.1 model to further enhance Grokโs capabilities on ๐. Premium and Premium+ subscribers are encouraged to update to the latest version of the ๐ app to participate in the Grok-2 beta test.
Whatโs Next?
Grok-2 and Grok-2 mini are being rolled out on ๐, with exciting applications in AI-driven features such as enhanced search capabilities, deeper insights on ๐ posts, and improved reply functionsโall powered by Grok. A preview of multimodal understanding as a core part of the Grok experience on ๐ and API will be released soon.
Since announcing Grok-1 in November 2023, xAI has been moving rapidly, driven by a small team with exceptional talent. The introduction of Grok-2 positions xAI at the forefront of AI development, with a focus on advancing core reasoning capabilities using a new compute cluster. xAI plans to share more developments in the coming months and is looking for individuals to join its dedicated team, committed to building the most impactful innovations for the future of humanity.
Read other articles: