Natural Language to Aliyun SLS SQL

Enter a natural-language query request. The server parses it into structured query JSON, validates fields and types, escapes SQL values, and renders an Aliyun Log Service SQL template.

Natural-language SLS SQL generator

Query request

Default detail limit

su match mode

SQL result

Generated SQL will appear here

Parsed JSON

Parsed JSON will appear here

Explanation

No explanation yet

Event dictionary

Event	Category	Description	action	actionValue

Field dictionary

Field	Type	Required	Description	Example

How to use this tool

Describe the query you want, such as filtering by su, grouping by type and trackType, or finding trackIds that contain event 6-1 but not 7-1, then generate SQL.

DeepSeek only parses natural language into ParsedQuery JSON. Field validation, SQL escaping, and template rendering stay in server-side deterministic code.

Example

For example, enter “query logs where su is https://www.fpffz.top/?channel=news01, grouped by type and trackType” to generate SLS SQL grouped by cc and ordered by cnt descending.

Frequently asked questions

Does DeepSeek generate SQL directly?

No. The model only returns ParsedQuery JSON, and final SQL is rendered by server-side templates.

Which query types are supported?

The generator supports group counts, raw filtered logs, records with A but without B, records with A but without B plus selected fields, and grouping by fields inside data JSON.

How is su matched by default?

su uses LIKE contains matching by default, which is useful for URLs with paths or query parameters. The tool also exposes exact matching.

Can I run the generated SQL without changes?

Use it as a starting point. Before production queries, verify field types, indexes, and SLS dialect functions for your project.

Privacy note

Your input is sent to the server-configured DeepSeek API for parsing. The API key is never exposed in the browser.