Elon Musk’s AI Company xAI Opens API Access to Grok 3

Even though Elon Musk is caught up in a legal battle with OpenAI, that hasn’t stopped him from pushing forward with his own AI company, xAI. In fact, Musk and his team just took a bold step — they’ve officially launched an API for their powerful AI model, Grok 3.

It’s been a few months since Grok 3 was introduced, and it’s clear this model is xAI’s answer to major players like GPT-4o from OpenAI and Gemini from Google. Grok 3 isn’t just another chatbot — it can understand and respond to images and questions, and it’s already powering features on Musk’s social platform, X (formerly Twitter). Not long ago, X actually acquired xAI, merging social media with cutting-edge AI tech in a move that raised more than a few eyebrows.

Now, developers and businesses can tap into Grok 3’s power through an API. There are two versions available: Grok 3 and a lighter version called Grok 3 Mini, both offering advanced reasoning abilities — a big deal for anyone working with AI.

But let’s talk price, because it’s not cheap. For Grok 3, it’s $3 per million input tokens (about 750,000 words) and $15 per million output tokens. The Mini version is more budget-friendly — $0.30 per million input tokens and $0.50 per million output tokens. If speed matters, you can pay more for faster access: up to $25 per million tokens for output.

Still, some folks aren’t too impressed. Grok 3’s pricing is on par with competitors like Claude 3.7 Sonnet by Anthropic, but more expensive than Google’s Gemini 2.5 Pro — a model that seems to outperform Grok 3 in many benchmark tests. There have also been whispers online that xAI’s benchmark claims for Grok 3 might be a bit too generous.

Users have also noticed that the API doesn’t support as much data in one go as promised. While xAI claimed Grok 3 could handle up to 1 million tokens, the actual API maxes out at 131,072 tokens (around 97,500 words). It’s a decent number, but still short of the grand claim made earlier.

When Musk first introduced Grok nearly two years ago, he painted it as a bold, rebellious kind of AI — not afraid to answer controversial questions or step outside the typical “safe” zone. And true to that image, early versions like Grok and Grok 2 were known for their uncensored, sometimes shocking responses. You could ask them anything, and they’d answer in ways that felt raw and unfiltered — something you wouldn’t see from ChatGPT.

But even Grok had its limits. When it came to sensitive political topics, the earlier models often played it safe or leaned noticeably to the left — especially on subjects like diversity, transgender rights, and social inequality. That surprised a lot of users who expected a more “neutral” voice. Musk pointed the finger at Grok’s training data — mostly public websites — and promised to steer future versions toward a more balanced tone.

Now, with Grok 3 out in the world, it’s still unclear whether that balance has truly been achieved. There have been a few awkward moments — like temporarily censoring mentions of Donald Trump and Musk himself — but we’ll have to wait and see how things evolve.

At the end of the day, Grok 3 is a reflection of Musk’s vision — bold, controversial, and still very much a work in progress. Whether it can truly compete with the likes of OpenAI and Google remains to be seen, but one thing’s for sure: it’s not afraid to make waves.