Grok is coming for the coding-focused prospects of its greater rivals. Elon Musk’s xAI dropped a brand new AI mannequin this week, particularly for builders, calling it a “speedy and economical reasoning mannequin” meant to assist with coding initiatives.
The brand new addition, dubbed grok-code-fast-1, helps spherical out the household of Grok 4 fashions, launched earlier this 12 months. The coding mannequin is agentic, which means it’s constructed to deal with duties and not using a ton of human oversight. xAI claims its mannequin outperformed OpenAI’s GPT-5, Anthropic’s Claude 4 and xAI’s personal Grok 4, saying it may remedy 70.8% of real-life software program issues, as evaluated by the SWE-bench.
Nonetheless, the model card notes that the coding mannequin has the next charge of dishonesty than Grok 4, as noticed by PCMag. So you will wish to double-check any essential work the AI does.
xAI co-founder Yuhuai Wu said on X that that is the primary of many coding fashions to return.
The beginning of a brand new mannequin line – Grok Code sequence.
This time we give attention to pace. It’s a really new expertise when the mannequin is tremendous quick.
Attempt it out, it’s free for 7 days on the foremost platforms. https://t.co/goJB8xCmPw— Yuhuai (Tony) Wu (@Yuhu_ai_) August 28, 2025
Agentic AI instruments have been a preferred growth within the ever-tightening AI race, with OpenAI, Anthropic and Microsoft every releasing a model of a coding agent. xAI’s device shall be extra much like Microsoft’s in that it is obtainable by way of its API, i.e., not open to paying ChatGPT or Claude subscribers.
You’ll be able to entry the brand new mannequin without cost for the following few days via quite a few unique platforms, together with Github Copilot, Cline, opencode, Cursor, Kilo Code, Roo Code and Windsurf. You will want an API key, which is priced out by utilization; you possibly can study extra about that with this breakdown of Grok models. xAI additionally put collectively a prompt engineering guide particularly for the coding mannequin to assist builders get probably the most out of it.
AI in coding initiatives generally is a tough conundrum. Generative AI instruments like chatbots have confirmed to be efficient instruments to assist inexperienced persons upskill shortly and assist professionals problem-solve. Agentic AI is starting to indicate the way it might help automate nonessential duties.
That poses the query of whether or not automated instruments may result in widespread employment disruption, significantly for younger software program engineers. A new Stanford study finds a excessive threat for staff early of their careers with important occupational publicity to AI, like software program engineering and customer support.
The common Grok chatbot and X/Twitter assistant has had its share of controversies. Earlier this summer season, the chatbot began producing extremely racist and antisemitic content, just a few days after xAI founder Elon Musk posted that the corporate had “improved Grok considerably.” xAI pulled the replace quickly after and shared that it had “taken motion to ban hate speech earlier than Grok posts on X.”
Bias is at all times a problem with AI chatbots, however Grok has illustrated how shortly the tech can spiral into getting used to shortly produce probably abusive content material.
(Disclosure: Ziff Davis, the mother or father firm of CNET, in April filed a lawsuit towards ChatGPT maker OpenAI, alleging it infringed Ziff Davis copyrights in coaching and working its AI techniques.)
