Which AI model should I be using right now (Claude vs Codex)?
The latest recommended model is Sonnet 4.5 (referred to on the call as 'Sonnet 5'), released the day before this call, which 'should be your go-to now.' There was also mention that 'Opus 5' (referred to as 'Fable 5') should be available again soon after previously being pulled. Codex was described as basically similar to Claude Code; one company reportedly splits usage 50/50 between Codex and Claude, with a personal preference for Claude due to familiarity.
This came up on one of our weekly calls.
Every Wednesday, members bring the real questions from their practice and work through them live with other accountants. Bring yours.
Apply to joinAlready a member? Sign in
More from the calls
If we want to build a smaller version of an internal AI/automation system, where should we start?
Jul 8 call
What should I expect if I want to sell my accounting/tax firm quickly, in terms of process and deal terms?
Jun 16 call
What backend tech stack is being used to build an internal AI-powered practice management tool (chat, transcripts, client data)?
Jul 8 call
How do I know when to move from a regular Claude chat to something like Claude Code or another Claude surface, and how do I avoid wasting tokens as a beginner?
Jul 1 call