xAI's Grok Code Fast 1 Accelerates Agentic Coding Speeds by Over 3x

San Francisco – xAI, Elon Musk's artificial intelligence company, has officially launched Grok Code Fast 1, a new reasoning model specifically engineered to enhance agentic coding workflows. Mohit Reddy, a key member of the xAI team, announced the release on social media, highlighting the model's development and the team's rapid progress. This release positions xAI as a significant contender in the competitive AI coding assistant market.

Grok Code Fast 1 is designed for speed and efficiency, delivering responses at approximately 160 tokens per second, which is more than three times faster than some leading competitors like OpenAI's GPT-5 (50.1 tokens/sec) and notably quicker than Gemini 2.5 Pro (92.4 tokens/sec) and Claude 4 Sonnet (78.7 tokens/sec). The model was built from scratch with a new architecture and trained on extensive programming content. It excels across the full software development stack, supporting languages such as TypeScript, Python, Java, Rust, C++, and Go.

The model's capabilities include building zero-to-one projects, providing insightful answers to codebase questions, and performing surgical bug fixes with minimal oversight. It has mastered common tools like grep, terminal, and file editing, making it highly compatible with integrated development environments (IDEs). On the SWE-Bench-Verified dataset, Grok Code Fast 1 achieved a 70.8% score, demonstrating its proficiency in real-world software engineering scenarios.

Initially released under the codename "Sonic," Grok Code Fast 1 is available for a limited free period through launch partners including GitHub Copilot, Cursor, Cline, Roo Code, Kilo Code, opencode, and Windsurf. Following this period, it will be priced competitively at $0.20 per million input tokens and $1.50 per million output tokens. This economical pricing strategy aims to make high-speed agentic development accessible to a broader range of developers.

xAI, founded by Elon Musk in July 2023, aims to "understand the true nature of the universe" and has rapidly assembled a "talent-dense team" comprising experts from Google, DeepMind, OpenAI, and Tesla. Co-founders Yuhuai (Tony) Wu and Jimmy Ba, both former students of AI pioneer Geoffrey Hinton, contribute deep expertise in machine learning and AI mathematics. Mohit Reddy's tweet emphasized the team's growth, stating, "What began as a two-person effort months ago is now a small, talent-dense team striving to improve and ship new model versions."

The company is actively recruiting experienced engineers and researchers to further its mission and accelerate the development of future AI models. Reddy concluded his announcement with a direct call to action for potential new hires, saying, "I believe we can transform how software engineers and organizations work in the coming years. Want to join us? Apply at https://t.co/wJ3wV3N8r7 or DM me directly." This ongoing recruitment drive underscores xAI's commitment to scaling its operations and expanding its impact on the AI landscape.