Published by xAI: "Grok Voice Agent API"
How media typically covers xAI
Based on 8 scored articles
xAI as author
“Published by xAI: "xAI Grok Imagine API Video Model Launch"”
Grok 4.1 achieves #1 ranking on LMArena's Text Arena with 1483 Elo and is preferred 64.78% of the time over previous Grok models, with improved emotional intelligence and reasoning capabilities.
“Published by xAI: "Grok 4.1"”
xAI's Grok 4 achieves state-of-the-art performance on frontier benchmarks including 15.9% on ARC-AGI V2 and 50% on Humanity's Last Exam, using a 200,000 GPU cluster and 6x improved compute efficiency through reinforcement learning at pretraining scale.
“Published by xAI: "Unpacking Musk's Bold AGI Comment"”
Grok 4 Fast achieves comparable performance to Grok 4 on frontier benchmarks while reducing token costs by 98% through 40% improved token efficiency and unified reasoning architecture.
“Published by xAI: "Grok 4 Fast Targets High Performance at Low Cost"”
Grok 4 Fast delivers frontier-level reasoning performance with 40% fewer thinking tokens and 98% lower cost than Grok 4, ranking #1 on LMArena's Search Arena.
“Published by xAI: "xAI Has a New Model - Grok 4 Fast"”
Grok 4 Fast achieves comparable performance to Grok 4 on benchmarks while using 40% fewer thinking tokens, resulting in 98% cost reduction and state-of-the-art price-to-intelligence ratio.
“Published by xAI: "Grok 4 Fast Is Here"”
Grok 4 Fast achieves comparable performance to Grok 4 on benchmarks while using 40% fewer thinking tokens and delivering 98% price reduction, setting a new frontier in cost-efficient reasoning with a 2M token context window.
“Published by xAI: "Grok 4 Fast"”
xAI launches grok-code-fast-1, a specialized coding model optimized for agentic workflows with exceptional speed and cost-effectiveness across multiple programming languages.
“Published by xAI: "Fast Models"”
xAI launches grok-code-fast-1, a specialized reasoning model optimized for agentic coding workflows that prioritizes speed and cost-efficiency, with launch partners including GitHub Copilot, Cursor, and others offering free limited-time access.
“Published by xAI: "xAI Releases Grok Code Fast 1"”
xAI open-sources Grok 2.5 model weights (~500GB) with instructions for deployment via SGLang inference engine requiring 8 GPUs with 40GB+ memory each.
“Published by xAI: "xAI Open-Sources Grok 2.5"”