Repository navigation
Go/Go Plus: per-model monthly limits read as separate buckets, but the console shows a single aggregate gauge #54263
Description
Summary
The Go / Go Plus pricing docs list per-model monthly limits (e.g. $60, $120, $180, $240) that read as separate buckets. But the console shows a single set of gauges (Rolling / Weekly / Monthly), and it is never stated whether the allowance is shared or per-model. The observed numbers only reconcile if the console gauge is an aggregate, not the max per-model — so the docs and the UI tell two different stories.
What the docs say
From https://opencode.ai/docs/go/ ("Usage limits"):
Each model has the following usage limits: 5-hour — 20% of the monthly limit; weekly — 50%; and monthly — 100%. Each model's monthly limit below determines how its usage counts toward those allowances.
…followed by a per-model table with a "Monthly limit" column ($60–$240 on Go Plus).
This is ambiguous: it can be read as (a) a single shared allowance into which each model's usage is normalized, or (b) a separate allowance per model. Nothing states which.
What the console shows
A single set of gauges for the whole account (Overview):
- Rolling (resets in 4h 40m): 1% used
- Weekly (resets in 1d 18h): 48% used
- Monthly (resets in 28d 3h): 24% used
- Cost
$24.83· Requests10,314· Tokens1.195B
Why it doesn't add up (as a user)
If the limits were separate per-model buckets, the "Monthly" gauge should show the max across models. In my case the model closest to its own limit is ~11–13%, yet the gauge shows 24%. That is only possible if the gauge aggregates (sums) the per-model consumption, or if there is a single pooled budget ($24.83 / 0.24 ≈ $103).
So the per-model dollar figures are misleading: they look like independent budgets, but the UI behaves as if they combine.
Request
- State explicitly in the docs whether the Go / Go Plus allowance is shared (one pool) or per-model.
- Explain what the per-model "Monthly limit" column means relative to the console gauge (a hard cap? a weight/normalizer?).
- Ideally surface it in the console too (e.g. a tooltip on the Monthly gauge).
Environment
- Plan: Go Plus
- Console → Overview (Rolling / Weekly / Monthly gauges)
- Docs: https://opencode.ai/docs/go/ (accessed 2026-10-10)
Activity
opencode-agent commentedon Oct 10, 2026
Thanks for the detailed write-up. You're right that the docs never say whether the Go / Go Plus allowance is one shared pool or separate per model.
I think this is a duplicate of #49186, which reports the same confusion (per-model limits that look separate but use a single shared quota) and asks for the same fix in the docs and console.
Possible duplicates and closely related issues, ordered from most to least likely:
The bot will close this issue as a duplicate in 1 day unless you reply explaining how your problem is different. Your concrete suggestions (an explicit statement in the docs and a tooltip on the Monthly gauge) are welcome on #49186 too.
redcasinotimonneo commentedon Oct 10, 2026
Hi there, please read my message about OpenCode usage limit: #49014 (comment)
Hope it helps @TempestiAlessio
opencode-agent commentedon Oct 11, 2026
We haven't heard back since the duplicate notice, so I'm closing this as a duplicate of #49186. That issue covers the same problem: the Go / Go Plus docs list per-model monthly limits that look like separate budgets, but the console counts usage against one shared Rolling / Weekly / Monthly allowance, and the docs don't say so.
Please follow #49186 for updates. Your suggestions would help there: say outright in the docs whether the allowance is shared, explain what the per-model "Monthly limit" column means, and add a tooltip to the Monthly gauge. #47547 and #49014 explain how the gauge adds up the per-model percentages. If your problem turns out to be different, reply here and we can reopen.