[FEATURE] Optional quota multipliers (x2 / x4 / x8 / x16) for Go subscription with volume discount

Feature Request: Optional quota multipliers (x2 / x4 / x8 / x16) for Go subscription

Context: how OpenCode limits actually work

Today the usage model is:

• Free tier: rolling windows (daily ~200 requests, plus 5-hour / weekly / monthly limits) — FreeUsageLimitError / 429 when exhausted.
• Go subscription: a flat quota that re-fills on fixed windows (5-hour rolling, weekly, monthly) — GoUsageLimitError when exhausted.
• Zen: pay-as-you-go credits used as fallback.

There is no "tier ladder" — Go is a single flat plan with one quota size. When a user's workload outgrows it (CI-heavy usage, parallel sub-agents, big refactors), the only options are:

1. Stay on Go and hit the quota mid-month → fall back to Zen pay-as-you-go, which is hard to budget and surprises at the end of the month.
2. Top up via Zen repeatedly — manual and easy to lose track of.

There is no way to "pay a bit more and get exactly 2× the same quota".

Proposed feature

Add an optional quota multiplier to the Go subscription: x1 (default), x2, x4, x8, x16. The multiplier scales all Go quota windows (5-hour rolling, weekly, monthly) by the same factor, and the monthly price scales accordingly — ideally with a modest volume discount at higher multipliers (e.g. x2 = 1.9×, x4 = 3.6×, x8 = 6.8×, x16 = 12.8× of the base Go price).

Key requirements:

• Choose the multiplier at checkout (e.g. "Go x4") and change it any time mid-cycle, prorated.
• Transparent pricing: show the scaled quota for each window (5h / weekly / monthly) and the exact monthly price before confirming.
• All quota windows scale together — no option to scale only one window.
• Optional and non-breaking: existing Go subscribers keep their x1 plan untouched.
• Cleaner than Zen fallback: the scaled quota re-fills on the same schedule as Go, so usage stays predictable and budgetable.

Use cases

• Solo developer: base Go quota is enough 80% of the time, but some weeks need 2× — switches to Go x2 for that period instead of paying Zen on top.
• Small team sharing one subscription: x4 covers a spike month without moving to a different billing model.
• CI / automation workloads: predictably high volume → x8 / x16 with a volume discount.

Why not just Zen?

Zen is pay-as-you-go — unpredictable monthly cost and no automatic quota re-fill. A multiplier keeps the "set it and forget it" subscription model while letting users right-size their quota.

Acceptance criteria

• Multiplier selectable at subscription time and changeable later (prorated).
• Price + scaled quotas (5h / weekly / monthly) shown for every multiplier before confirmation.
• opencode usage (see #32913) reflects the scaled quotas.
• Volume discount documented and applied at higher multipliers.

Related

• #24879 (Pro tier proposal) — overlaps in the "more quota" goal but proposes a separate tier; a multiplier is orthogonal and can complement it.
• #32913 (CLI usage view) — would display the scaled limits.

Happy to clarify or help test if this gets picked up.