Kol Tregaskes @koltregaskes

Peter compares ChatGPT (GPT-5), Claude (4.5), and Gemini (2.5) across 10 tasks, recommending the best model for each based on personal use.

- Prefers ChatGPT for everyday answers due to speed, conciseness, and fewer rate limits compared to Claude.
- Chooses Claude for writing and editing, trained on personal style examples to match output.
- Recommends Claude Code or OpenAI Codex for coding, with Codex better at fixing complex bugs.
- Selects Gemini for web search for faster real-time info retrieval.
- Favours Claude for deep research to generate concise five-page reports.
- Picks ChatGPT for voice chat despite interruptions, prefers third-party apps like Wispr Flow for dictation.
- Recommends Gemini's Nano Banana for image generation and Veo 3.1 for video, but uses Sora for personal videos.
- Finds no model useful for live camera or computer use.
- Dismisses Grok due to its AI companions marketing, seen as unprofessional for work.
- Highlights AI Projects, preferring Claude for attaching plain text files.
- Shares tips: attach Google Workspace files, specify output styles, iterate prompts.
- Details top three projects for product work: edit product reviews/PRDs, draft performance reviews, write interview summaries.
- For creator work: prep podcast interviews, draft post-production, edit newsletter posts.
- For personal work: get life advice, plan weekends/trips, home improvement tips.
- Describes Deep Research as preferred for complex problems, with five steps to optimise (details not fully provided).

Peter provides practical workflows and preferences to maximise AI for daily tasks.

Nov 2, 2025 View on X →

https://x.com/koltregaskes/status/1985051785153626264