xAI releases Grok-2 and Grok-2 mini in beta

xAI/Grok

xAI released Grok-2 and Grok-2 mini to X Premium subscribers, with an enterprise API to follow in August. An early anonymous test on LMSYS ranked Grok-2 above Claude 3.5 Sonnet and GPT-4 Turbo; the release added image generation via Black Forest Labs' FLUX.1 model.

• Grok-2 achieves 56.0% on GPQA (graduate-level science) and 87.5% on MMLU benchmarks.
• Grok-2 mini offers a balance between speed and answer quality for faster responses.
• State-of-the-art results in visual math reasoning and document-based question answering.
• Enterprise API to feature multi-region inference deployments and mandatory multi-factor authentication.