# 按计算查询成本付费

By [ma1984.eth](https://paragraph.com/@ma1984) · 2024-01-29

---

价钱 当您使用空间和时间时，您所付出的只是计算。存储、索引区块链数据、OpenAI 仪表板、集群中的 OLTP + OLAP 查询、SQL 加密证明等始终包含在内。

我们提供多种不同的定价选项来满足您的需求和用例。请访问我们的定价页面，详细了解哪种选项最适合您。

关于按计算付费 按计算付费是 Space and Time 的基于使用情况的定价选项，让您只需为运行的查询付费。计算成本是使用类似gas的模型来计算的，其中计算消耗以空间和时间积分来衡量。消耗的计算量取决于确切的查询、数据大小以及返回的数据量。

为了让您了解在按计算付费模型下不同查询的成本，我们提供了一组针对各种 public\_read 表的查询及其经过测试的积分消耗。

目前，1 个计算积分的成本= 1 USD。

🚧 关于测试版的注意事项

请注意，按计算付费仍处于测试阶段，下面的估算可能会随着时间的推移而发生变化。您可以随时返回此页面查看最新的估算。

一般观察：大型表按字段进行分区，这意味着只要您对未修改的字段进行操作，时间限制查询将比无限制查询（也称为全表扫描）消耗更少的资源，如下面的示例所示。 Time\_StampTime\_Stamp

如果您需要对所有无限历史记录（即获取地址的完整历史记录）进行频繁查询，您可以考虑创建物化视图。

单行表查找 SQL

SELECT \* FROM sxtlabs.singularity 挂钟时间 制作人员 93毫秒 0.000037125 多行查找 SQL

SELECT \* FROM sxtdemo.stocks WHERE STOCK\_DATE = '2023-11-21' 挂钟时间 制作人员 94毫秒 0.000621000 简单解析 SQL

SELECT SYMBOL, AVG((STOCK\_HIGH-STOCK\_LOW) / STOCK\_OPEN) AS Avg\_Volatility FROM sxtdemo.stocks WHERE SYMBOL in ('MSFT','AMZN','GOOGL') GROUP BY SYMBOL 挂钟时间 制作人员 748毫秒 0.008040050 简单分析，大扫描 SQL

SELECT cast(TIME\_STAMP AS date) AS Create\_Date, count(\*) AS Contracts\_Created FROM POLYGON.CONTRACTS WHERE TIME\_STAMP BETWEEN '2023-01-01' AND '2023-04-30' -- 200k of 2M ROWS, 10% GROUP BY 1 ORDER BY 1 desc 挂钟时间 制作人员 5秒 0.222452000 中等分析 SQL

SELECT avg(Block\_Count) AS DailyAvg\_Block\_Count ,avg(Validator\_Reward\_Matic) AS DailyAvg\_Validator\_Reward\_Matic ,avg(Gas\_Used\_Matic) AS DailyAvg\_Gas\_Used\_Matic ,round(avg(Avg\_Txn\_per\_Block),2) AS DailyAvg\_Txn\_per\_Block FROM ( SELECT cast(TIME\_STAMP AS date) AS Block\_Date ,sum(REWARD)/1e18 AS Validator\_Reward\_Matic ,sum(GAS\_USED)/1e18 AS Gas\_Used\_Matic ,AVG(TRANSACTION\_COUNT) AS Avg\_Txn\_per\_Block ,count(\*) AS Block\_Count FROM POLYGON.BLOCKS WHERE TIME\_STAMP BETWEEN '2023-01-01' AND '2023-01-26' -- 1M ROWS GROUP BY 1 ) AS a 挂钟时间 制作人员 2秒 0.058209900 大型分析 跨 3 个不同的链查询整个 EVENT.LOG 表（曾经触发的每个智能合约事件 - 任何模型中最大的单个表）。

SQL

SELECT p.Event\_Date ,p.Event\_Cnt AS Polygon\_Event\_Count ,p.Event\_Cnt - lead(p.Event\_Cnt) over(ORDER BY p.Event\_Date desc) AS Polygon\_Event\_Growth ,round((1-(lead(p.Event\_Cnt)over(ORDER BY p.Event\_Date desc) / p.Event\_Cnt))_100,2)||'%' AS Polygon\_Event\_Growth\_Pct ,e.Event\_Cnt AS Ethereum\_Event\_Count ,e.Event\_Cnt - lead(e.Event\_Cnt) over(ORDER BY e.Event\_Date desc) AS Ethereum\_Event\_Growth ,round((1-(lead(e.Event\_Cnt)over(ORDER BY e.Event\_Date desc) / e.Event\_Cnt))100,2)||'%' AS Ethereum\_Event\_Growth\_Pct ,s.Event\_Cnt AS Sui\_Event\_Count ,s.Event\_Cnt - lead(s.Event\_Cnt) over(ORDER BY s.Event\_Date desc) AS Sui\_Event\_Growth ,round((1-(lead(s.Event\_Cnt)over(ORDER BY s.Event\_Date desc) / s.Event\_Cnt))100,2)||'%' AS Sui\_Event\_Growth\_Pct FROM ( SELECT CAST(p1.TIME\_STAMP AS date) AS Event\_Date ,count(_) AS Event\_Cnt FROM POLYGON.LOGS AS p1 WHERE p1.TIME\_STAMP BETWEEN '2023-08-01' AND '2023-08-31T23:59:59' GROUP BY 1 ) AS p LEFT OUTER JOIN ( SELECT CAST(e1.TIME\_STAMP AS date) AS Event\_Date ,count() AS Event\_Cnt FROM ETHEREUM.LOGS e1 WHERE e1.TIME\_STAMP BETWEEN '2023-08-01' AND '2023-08-31T23:59:59' GROUP BY 1 ) AS e ON p.Event\_Date = e.Event\_Date LEFT OUTER JOIN ( SELECT CAST(s1.TIME\_STAMP AS date) AS Event\_Date ,count() AS Event\_Cnt FROM SUI.EVENTS AS s1 WHERE s1.TIME\_STAMP BETWEEN '2023-08-01' AND '2023-08-31T23:59:59' GROUP BY 1 ) AS s ON p.Event\_Date = s.Event\_Date 挂钟时间 制作人员 10秒 0.52894240

---

*Originally published on [ma1984.eth](https://paragraph.com/@ma1984/F9o8dFRvo65go5Cov8Fd)*