OpenCode Go plan usage estimate: see all models on one page

Ever since I signed up for the OpenCode Go plan, one thing has been nagging at me: how many requests can I actually use each month?

This thing kept shifting all through August — prices for models at every tier were being adjusted, new models kept getting added, plus assorted promos. Basically I had to hit the official site every single day to see which models my plan’s quota covered and how many runs each one got. Exhausting.

The kicker is that the monthly quota data on the official site is scattered across several different tables:

• The pricing table (unit price per 1M tokens) is one table
• The monthly usage quota ($15/$30/$60 tiers) is mixed into the pricing table
• The per-request token composition (input/cache/output) is yet another table
• And the official per-model “requests per month” numbers need a separate table of their own

Meanwhile that “5-hour quota” chart everyone loves to look at doesn’t list all the models either — you only get a rough idea.

Here’s what it looks like:

The Approach

So I handed the whole problem to an agent.

Once ds picked up the task, it reasoned: if the official “requests per month” table is just back-derived from unit price × per-request token composition, then I can pull the latest pricing table myself, run it through the same formula, and get the complete usage figures for every model. Then chart it and take it all in at a glance.

The formula is simple:

每请求成本 = (输入token×输入价 + 缓存token×缓存价 + 输出token×输出价) / 1,000,000
每月请求数 = 每月使用额度 ÷ 每请求成本

This lines up with the logic the official site uses to generate its “request estimate table.” In testing, most models come out within <1% of the official values, which effectively validates the formula.

Implementation

The whole tool is a single dependency-free HTML file that fetches data automatically when you open the page:

• The data source is the raw mdx on the dev branch of the GitHub repo anomalyco/opencode (kept in sync with the opencode.ai docs), with a jsdelivr CDN source as a fallback. Both send CORS headers, so the browser can fetch cross-origin directly — no backend needed.
• It parses three things: the price + quota table, the request pattern table (per-request token composition), and the official request count table (used for comparison).
• Model names are normalized with longest-prefix matching, to handle the official site’s combined notation like GLM-5.3/5.2/5.1 and Kimi K2.7 Code.
• Then it draws a horizontal bar chart (log scale, so the bigger the usage the longer the bar). Hovering any bar shows the details: unit price, monthly quota, cost per request, token pattern, and the official comparison value.
• If a fetch fails, it automatically falls back to the last successful localStorage cache, and there’s a “Refresh now” button at the top.

Bars with a discrepancy get a ▲ marker. For multi-tier models (like DeepSeek V4 Flash’s Peak/Off-Peak), each tier gets its own bar, while the official site only gives one aggregated figure, so it usually won’t match the higher tier — that’s by design, not a miscalculation.

The Result

Now when I want to check on my plan any day, opening this one page is enough — no more flipping back and forth between several tables on the official site.

Live preview: https://mousebomb.org/opencode-go-limits/

If you’re on OpenCode Go too, this page should save you some of that daily table-hunting.