Another day, another model - Grok 3 Launch

This week Elon's xAI released their Grok 3 model with claims that it beats the big players such as OpenAI and DeepSeek.

Currently there is no universally agreed upon and standardised way to test and compare AI models, it's difficult to substantiate these claims. However at first glance is seems quite good.

For most of us, this news doesn't mean much other than there being another competitor in the game which will keep the AI market competitive for the rest of us.

Key summary:

• Musk claims Grok 3 to be "maximally truth-seeking AI", presumably meaning it has less politically correct filters and bias
• The model is available for free in a suspiciously similar UI to ChatGPT at grok.com
• A pro version is available for USD $30/month which unlocks higher limits, DeepSearch, unlimited image generation and Grok 3 "thinking", all very similar to what's available with ChatGPT
  • With the exception of DeepSearch, as ChatGPT's "deep research" equivalent is still only available to USD $200/month "pro" users

My 2 Cents on the Launch of Grok 3

It's always good to have another competitor in the game keeping OpenAI on their toes. I also like the idea of model potentially having less bias and filtering on it, although we're yet to see what that actually means given the current political climate.

Ultimately AI models, whether through a ChatGPT-style UI or API are becoming rapidly commoditized and to the average user it's much of a muchness.

Personally I won't be using Grok 3 as I can get everything I need from ChatGPT, Perplexity and Cursor (coding).