Developer ditches ChatGPT desktop for T3 Code after third try, citing frictionless multi-model switching
aniketapanjwani · x · 2026-11-15
Developer Aniket Panjwani explains why, after two failed attempts over 6 months, he finally moved all his work from the ChatGPT desktop app to Theo's T3 Code.
Highlights:
- Frictionless switching between models (his default: Opus 5.5 high), which makes him more inclined to add Gemini and Grok subscriptions
- Chats get "settled" instead of archived and stay visible in the sidebar
- Easier mobile-to-desktop pairing than ChatGPT or Claude desktop apps
- Well-designed token usage/limits UI across all providers, richer annotation tools (draw/region), and Computer Use via OpenAI models
Papercuts: no text-selection annotation, no vim-style annotation mode that aggregates marks like ChatGPT desktop, and no in-app camera capture on mobile.
He was planning exhaustive ADE comparisons but is satisfied enough with T3 Code to stick with it.
More from coding & agent
- DeepMind's AlphaProof Code formally verifies Signal crypto with 6M lines of Lean in 4 weeks — dl_weekly · 2026-11-15
- 16x Microsoft MVP builds native PowerShell agent on Microsoft's Intelligent Terminal — dfinke · 2026-11-15
- Raindrop lets coding agents create accounts before humans claim them, with write-only keys — New-Tailor4071 · 2026-11-15
- Write the eval runner before the agent: evals-first architecture with Ports and mock adapters — hwchase17 · 2026-11-15
- How do you evaluate agents whose skill library keeps evolving between test runs? — tahaan · 2026-11-15
- One lint fix burned 50M tokens: agent plugin nukes Codex Pro quota in 12 hours — generativist · 2026-11-15