Hey @zhaoxi2000,
You’re right: on plans with legacy request-based pricing, Gemini 4 Argon now requires Max Mode, the same as the other frontier models (GPT-5.x, Opus, and so on). Since the release of GPT-5.4, new frontier models have required Max Mode on legacy request-based plans. There’s more background in this thread. On usage-based pricing, Max Mode doesn’t apply and nothing changes.
To keep using Argon, select it and Max Mode turns on automatically. On legacy plans, Max Mode is billed at the model’s API rate plus 20% (Max Mode on legacy plans). If you’d rather stay on flat per-request billing, models that don’t require Max Mode still work without it; the models page lists which ones do.