Updated 11.4.2024: Added second day case too without using fast mode.

There has been a lot of brouhaha in e.g. Reddit about recent (10.4.2026) changes to plans: OpenAI announced new 100€ ‘pro lite’ tier, and there was worry that ‘plus’ tier was cannibalized for it. So I decided to use my 5 hour quota using fast mode gpt-5.4 (high for planning, medium for implementation), and then subsequently second time without fast mode, same model and thinking parameters.

How much did I get out of it?

┌──────────────┬──────────────────────────────────────────────────────┬──────────────┬────────────┬────────────┬────────────────┬────────────────┬─────────────┐
│ Date         │ Models                                               │        Input │     Output │  Reasoning │     Cache Read │   Total Tokens │  Cost (USD) │
├──────────────┼──────────────────────────────────────────────────────┼──────────────┼────────────┼────────────┼────────────────┼────────────────┼─────────────┤
│ Apr 10, 2026 │ - gpt-5.4                                            │    1,069,508 │     94,045 │     38,453 │     13,131,648 │     14,295,201 │       $7.37 │
├──────────────┼──────────────────────────────────────────────────────┼──────────────┼────────────┼────────────┼────────────────┼────────────────┼─────────────┤
│ Apr 11, 2026 │ - gpt-5.4                                            │    1,752,868 │    164,757 │     54,636 │     31,659,392 │     33,577,017 │      $14.77 │
└──────────────┴──────────────────────────────────────────────────────┴──────────────┴────────────┴────────────┴────────────────┴────────────────┴─────────────┘

Even the first attempt (one 5h window), in the fast mode, produced quite a lot. This was 91% of my 5 hour quota, and 14% of weekly limit, and it is already 7$ in API token costs. 14 million (mostly cached tokens) for (roughly) 2,8% of my monthly quota, which means that this essentially cost me perhaps 0,7€ or so.

The second day (another 5h window) was done without fast mode, using 100% of 5 hour quota (+- rounding error, UI shows 0% left) and 16% of weekly limit. It proved ccusage does not support fast mode, as the price estimate jumped to double ( as I used twice as many tokens presumably ).

Thoughts

Either

  • this sample is too small, or
  • the amount of tokens you get is perhaps halved for the same subscription (based on the previous monthly analysis of late February / most of March), although they had 2x promotion going at the time and also occasional quota resets.

Based on the second sample, if I have more time for development, I can apparently exhaust the Plus plan 5 hour quota in about an hour and half without using fast mode, with 2 worktrees churning away. So even single worktree cannot be fully employed with how I use it for full 5 hours, but still, I seldom code ‘all the time’, at least hobby stuff. If I use fast mode on the other hand, it seems I can exhaust Plus 5h quota right now with 1 worktree in 1 hours, getting 1.5 hours’ effort out of it (which is fair trade, if I am planning to do only little).

Given typical 1:10 noncached:cached pricing structure, paying 23€ monthly for this is still crazily good deal as of now. Especially when I am not actively developing all the time, fast mode Codex seems to be still the best deal in town (I am sort of missing Spark, which I no longer have access to - it was very nice for fast and simple things, even if it got things wrong often).