Does Fast Mode mode work through Codex LB?

[josevelaz]

Ive enabled fast mode in opencode but im not seeing any difference in speed or quota usage.

To enable fast mode in opencode:

1. Overload the openai/gpt-5.4 provider/model
2. add an options object
3. set "serviceTier": "priority"