xAIlaunched its Grok-3 AI chatbotmerely a few days ago , butlocked it behind a paywall worth $ 40 per month . Now , the company is offering liberal entree to it , but only for a limited time . xAI chief , Elon Musk , tell the spare access will only be available for a “ unretentive metre , ” so it ’s anyone ’s guess how long that windowpane is going to be .
For now , the only two features uncommitted to meet around are Think and DeepSearch . Think is the lineament that adds conclude capabilities to Grok-3 interactions , in the same view asDeepThink on DeepSeek , Google ’s Gemini 2.0 Flash Thinking Experimental , andOpenAI ’s type O - series models .
Thinking and reasoning model show their train of cerebration on how they develop down and eventually sue the user enquiry . The outcome , as per experts , is better carrying out at tasks such as solve scientific discipline , coding , and mathematical problem .
For a unretentive time , Grok 3 is available for free to all!https://t.co / r5iLXi2pBm
& mdash ; Elon Musk ( @elonmusk)February 20 , 2025
DeepSearch , on the other mitt , is xAI ’s tantamount ofDeep Research tool that is now available for Perplexity , Gemini , and ChatGPT user . Grok-3 is only the 2nd mainstream AI product out there volunteer spare admittance to a compute - intensive outgrowth such as DeepSearch or Deep Research .
This is one of the most promising agentic manipulation cases of an AI simulation , as it takes the cognition foregather process to a whole new story . Once users advertize their question , it is break down and a inquiry plan is presented , include details of what source the solution must be extract from .
The AI goes though all the relevant repositories of knowledge , reasons through the datum compile in real - time , and presents it in the form of a comprehensive written report .
If you want to access Grok-3 on the wandering platform , head over to the X app and tap on the Grok icon in the bottom bar . On the web , it is accessible directly via the social medium platform by visiting thex.com/i/grokpage .
There is substance to all the hype
In my brief time with DeepSearch , I have found it to be quite telling . Unlike Gemini Deep Research , it does n’t ask for an approval of the research methodology and forthwith goes to project once you submit your motion . Compared to Gemini , it ’s also quicker .
Regarding a deep inquiry inquiry about the scientific research situation on the wallop of silver screen sentence on young psyche , Grok-3 provided a write up quicker than Gemini . Moreover , it is more transparent , as you may see the elaborate intellection process that lead behind pick up and finding the solvent .
Moreover , you’re able to see a stage - broad breakdown of the whole mental process , advance in real - clock time and bring through as its own dataset , above the real answer . On the flip side , you could not edit the research design before the agentic search begins .
Gemini , on the other handwriting , take a more opaque overture . you’re able to edit out the research plan , but you ca n’t see a process breakdown or the thought - reasoning flow . As far as caliber goes , Gemini cite material from 37 website , and provided them all as a footer for quote . Grok-3 ’s DeepSearch only list six key citations , even though the solvent it provided were no less utile .
When I essay the less intensive Think lookup , Grok-3 proved to be the quicker among the two , once again . I ask about the relevancy ofMicrosoft ’s new quantum computation chip , and noticed two all important remainder .
Grok-3 take a more holistic approach to answering it , focusing not only on the scientific software and the benefits , but also the risks that come with it , in a dedication department of its own . Another difference is that you’re able to see the chain of abstract thought at any give fourth dimension .
Gemini 2.0 Flash Thinking Experimental is , once again , opaque , though its section - wise breakdown was more comprehensive . Another crucial difference is that Gemini ’s answers appear more reticent and academically - inclined , while Grok-3 takes a more libertarian attack at explicate complex terms .
It ’s voiceless to declare which AI model is superior , but according to benchmarks shared by xAI , Grok-3 has top Google , OpenAI , DeepSeek , and Anthropic ’s AI models on multiple evaluation charts .
