TSX lower after index logs fresh record closing high
Investing.com -- xAI has introduced Grok Code Fast 1, a new AI model specifically designed for coding tasks that prioritizes speed and cost-effectiveness.
The model features a brand-new architecture built from scratch, with a pre-training corpus rich in programming content and post-training on datasets reflecting real-world coding tasks.
xAI has partnered with several platforms to offer the model for free for a limited time, including GitHub Copilot, Cursor, Cline, Roo Code, Kilo Code, opencode, and Windsurf.
Grok Code Fast 1 is particularly proficient in TypeScript, Python, Java, Rust, C++, and Go. It can handle various programming tasks from building new projects to fixing bugs with minimal supervision.
The model is priced at $0.20 per million input tokens, $1.50 per million output tokens, and $0.02 per million cached input tokens, making it an economical option for developers.
On the SWE-Bench-Verified benchmark, Grok Code Fast 1 achieved a score of 70.8% using xAI’s internal testing harness. The company noted that while benchmarks provide valuable insights, they supplemented these with human assessments from experienced developers to evaluate real-world performance.
The model was previously released in stealth mode under the codename "sonic" last week, allowing the team to gather feedback and deploy multiple new model checkpoints.
xAI has stated they plan to deliver consistent updates to Grok Code Fast 1 with improvements arriving in days rather than weeks. A new variant supporting multimodal inputs, parallel tool calling, and extended context length is already in development.
This article was generated with the support of AI and reviewed by an editor. For more information see our T&C.