After being a power user of AI for years, this is the model I recommend for each use case (Dec 2025): • Coding/apps/dashboards - Claude Opus 4.5 • General chatting/voice prompting/strategy - GPT 5.1 (still the best voice prompting by a mile, which i really value) • News/finance/real-time info - SuperGrok (i'm an X power user so I find it really helpful) • Image gen - Gemini 3 (I also use Gemini for other niche use cases like YT retention/writing - but usually mixed with other LLMs - I'd honestly make Gemini my #1 if it weren't for the awful voice transcription which keeps me hooked to GPT) • Deep research (for investments, content etc.) - a mix of Perplexity Pro/5.1 Pro/Claude Research I'll reply to as many comments I can if you'd like to know my preferred model for niche use cases(Miles Deutscher)