Grok review
Live-information assistant from SpaceX's AI unit, with X-native search and multi-agent reasoning.
TECHi verdict
Grok is the best assistant for one question shape — "what is happening right now, and what does it mean" — and a middling choice for most other work. SuperGrok at $30 is worth it if X is your information diet: the live grounding is real, the multi-agent mode is genuinely novel, and the model cadence since the SpaceX merger has been fast. Heavy at $300 is for people who need 4.5's depth professionally, not enthusiasts. For general writing, coding, and analysis, ChatGPT and Claude remain stronger daily drivers. The open question is stewardship: a rocket company's AI unit training frontier models on Cursor data is a fascinating machine, but nobody buying a consumer chatbot subscription can be sure where its priorities will sit in a year.
Pros
- + Live web and X grounding by default — the strongest real-time answers of any major assistant
- + Auditable multi-agent mode shows each agent's reasoning instead of a single black-box answer
- + Fast model cadence: Grok 4.5, a 1.5T-parameter frontier model, shipped publicly in July with aggressive API pricing
Watchouts
- - Answer quality leans on X's firehose, so hot-topic responses need more source-checking than Perplexity or ChatGPT Search
- - Frontier access is expensive — full Grok 4.5 sits behind the $300/month Heavy tier
Frequently asked
What are the pros of Grok?
Live web and X grounding by default — the strongest real-time answers of any major assistant. Auditable multi-agent mode shows each agent's reasoning instead of a single black-box answer. Fast model cadence: Grok 4.5, a 1.5T-parameter frontier model, shipped publicly in July with aggressive API pricing
What are the watchouts / cons of Grok?
Answer quality leans on X's firehose, so hot-topic responses need more source-checking than Perplexity or ChatGPT Search. Frontier access is expensive — full Grok 4.5 sits behind the $300/month Heavy tier
What is the TECHi verdict on Grok?
Grok is the best assistant for one question shape — "what is happening right now, and what does it mean" — and a middling choice for most other work. SuperGrok at $30 is worth it if X is your information diet: the live grounding is real, the multi-agent mode is genuinely novel, and the model cadence since the SpaceX merger has been fast. Heavy at $300 is for people who need 4.5's depth professionally, not enthusiasts. For general writing, coding, and analysis, ChatGPT and Claude remain stronger daily drivers. The open question is stewardship: a rocket company's AI unit training frontier models on Cursor data is a fascinating machine, but nobody buying a consumer chatbot subscription can be sure where its priorities will sit in a year.