Is anybody having Codex repeatedly and silently downgrade the model from GPT 5.6

<!-- SC_OFF --><div class="md"><p>The recent changes to Codex limits have significantly degraded the developer experience.</p> <p>Previously, Codex in VS Code felt fluid and

<!-- SC_OFF --><div class="md"><p>I've been working all day on ONE THING. One function. Can't get it right. I've been working all week in Codex and it's been generally fine, but

<!-- SC_OFF --><div class="md"><p>I've been using Codex (the Desktop program) and I'm having weird problems. In some conversations it starts working in roughly 1 minute

<!-- SC_OFF --><div class="md"><p>For myslef im using it a lot but it will produce so much garbage its a full time job it feels like just maintaining scope. I combined all the

<!-- SC_OFF --><div class="md"><p>New Codex limits are a joke. They said they would cut the 2x usage bonus and cut more of the 5hours limits, but the consumption has raised to

<!-- SC_OFF --><div class="md"><p>As of today I can't find my project folders on the left pane above recents anymore - I see this spaces stuff and there's a folders open there but

<!-- SC_OFF --><div class="md"><p>I gave my Codex agent a task that would normally take me at least 12-15 hours — multiple steps, logic handling, and some debugging involved.<br

<!-- SC_OFF --><div class="md"><p>I'm reporting this for the updated Aplha 2 update version 0.124.</p> <p>Was scheduled to perform 4 NIAH tests with a local model after being