xAI is pushing Grok Build to v0.2.11! Poaching two core members from Cursor in an attempt to catch up with Claude Code

📄Full Article· Automatically extracted by trafilaturaGemini 翻譯2208 words

In less than half a month, xAI has pushed its AI coding tool Grok Build from v0.1 to v0.2.11, adding features such as 8-agent parallel processing, X platform search integration, and Plan Mode, in an attempt to capture market share in the AI coding space. (Context: Building a $3 trillion super-beast? Rumors suggest Musk intends to merge Tesla and SpaceX; if successful, the new entity would hold 30,000 BTC.) (Background: Altman retracts his "AI will destroy human jobs" prediction: "I'm glad I was wrong, truly.") Key Highlights - xAI's Grok Build went from its May 14 launch to v0.2.11 in half a month, supporting 8 sub-agents in parallel, X search integration, and a default-enabled Plan Mode. - xAI poached engineering and product leads from Cursor and secured a $60 billion acquisition option for Cursor's parent company, with a $10 billion breakup fee if abandoned. - SWE-bench Verified benchmark score is 70.8%, still showing a significant gap compared to Claude Code (87.6%) and OpenAI Codex CLI (85-88.7%). With 11 versions released in half a month, xAI's AI coding tool Grok Build has covered the update cadence of most competitors over six months in just two weeks. This CLI-first Agentic Coding tool launched in early beta on May 14, powered by the proprietary grok-code-fast-1 model with a 256K token context window, supporting up to 8 sub-agents operating in parallel within independent Git worktrees. The built-in Plan Mode is enabled by default (requiring user approval before editing code) and is currently the only AI coding tool that integrates real-time X platform search, allowing users to query community discussions and track API controversies directly within the development workflow. The latest round of updates focuses on four areas: In terms of agent capabilities, sub-agents can now share terminal backends and task schedules across sessions, with the addition of a "laziness detector" and proactive reminder mechanism. Regarding compatibility, support for Windows ARM64 and macOS x86_64 has been added, and issues with Linux Wayland and WSL copy-paste have been fixed. In terms of user experience, terminal video playback has been boosted to 30FPS, with support for multi-image pasting and macOS screenshot shortcuts. Regarding stability, the default retry budget has been increased to approximately 5 minutes, strengthening timeout handling for background tools. Poaching Two Key Figures from Cursor Beyond just sprinting through version numbers, xAI is simultaneously targeting talent. They poached Engineering Lead Andrew Milich and Product Lead Jason Ginsberg from the $50 billion-valued Cursor parent company, Anysphere, and brought in Devendra Singh Chaplot from Mistral AI. More notably, xAI signed an option agreement with Anysphere in April, allowing them to acquire the entire Cursor team and product for $60 billion later in 2026, with a $10 billion penalty if they walk away. Cursor currently uses xAI's Colossus supercomputer in Tennessee to train its models. Musk admitted in March that "xAI didn't get it right the first time and is rebuilding from the foundation," and all 11 founding members of xAI have since departed. SWE-bench Gap and AI Coding Goals Grok Build scored 70.8% on SWE-bench Verified, while Anthropic's Claude Code reached 87.6% and OpenAI Codex CLI reached 85-88.7%, a gap of nearly a full generation. However, xAI's strategy goes beyond just chasing benchmark scores; its 8-agent parallel processing is the highest among similar tools, and Grok Build can directly read Claude Code users' skills, plugins, and hooks configuration files, enabling zero-cost migration. A Medium tester discovered that Grok Build automatically loaded 47 skills upon startup, 12 of which came from the Claude Code configuration directory. The AI coding market is expected to reach $12.8 billion by 2026, with 84% of developers using or planning to use AI tools, and a median of 3.1 tools used per person. Cursor leads with $2 billion in annual revenue and 1 million paid users; Claude Code has captured 28% of the primary tool market share and was voted the "#1 Favorite Tool" by 46% of users; GitHub Copilot has 4.7 million paid subscribers. Grok Build pricing follows xAI's subscription model: SuperGrok ($30/month), X Premium+ ($40/month), and

Data Status✓ Full text extractedRead Original (動區 BlockTempo)

🔍Historical Similar Events· Keyword + Asset Matching6 items

2026-05-28

Anthropic launches its latest model, Claude Opus 4.8! Claude Code simultaneously adds "Dynamic Workflows" — one person can match a hundred-person development team

Similarity 120%關鍵字 code同分類 zh

2026-05-26

Two months after Claude Code burned through Uber's annual budget, the COO bluntly stated: There is no proportional relationship between Token consumption and useful output.

Similarity 120%關鍵字 code同分類 zh

2026-05-24

My 30-person company went all-in on Claude Code, and the result is "humans have more work to do than ever."

Similarity 120%關鍵字 code同分類 zh

2026-05-24