Here is when to use each top AI model right now š”
ā ā ā
ChatGPT 4o - best ideas/invention but follows directions worst. I have screamed at the new 4o multiple times. Use this for first draft. Better at organization. GPTs still the best prompt efficiency hack. Stack them. Voice dictate whenever possible
OpenAI o3-mini-high - tied for best with coding, amazing price (1/3rd price of o1 mini)
o3 / Deep Research - use for any āI wonder...ā, kick it off at the beginning of team meetings, always ask for exec summary in shorthand, better than Googleās Deep Research but still hallucinates! If you can afford it ($200/mo), likely the better option for business use
Claude 3.5 Sonnet - most creative writing while maintaining my personal writing style, most helpful, āgets meā the most, tied for best with coding, use this to fix writing style
Grok 3.0 - most inventive writing but doesnāt maintain style unless you prompt the heck out of it, most fun, best image generator (yes, you heard that right), very millennial-sounding, has its own style, use sparingly for viral soundbites, up there for coding but low rate limit so not first pick
Gemini 1.5 Pro - best recap, āgets meā second most often after Claude, most balanced (I use it because it takes in video and 2.0 does not), I love to use this at the end of a project where I screenrecord entire AI conversations with other models and then ask Gemini to weigh in
Google Deep Research - 1/10th the cost of OpenAI Deep Research and a close second place, most people should use this one over OpenAI for personal use bc of cost-performance tradeoff
Web-browsing agents (Operator, Proxy, others in alpha) - I donāt like any public ones enough yet
DeepSeek - honestly, other than the natural CoT, this isn't in my top as a user. As a builder/investor, of course v interesting
ā ā ā
šØ This is only my opinion. Your AI mileage may vary. And I fully agree with Ethan's sentiment below.
ā»ļø Be sure to like, share, and save this tweet.