diff --git a/dumpling-overview.md b/dumpling-overview.md
index 4aa03bfe8c149..79cb186dbacf7 100644
--- a/dumpling-overview.md
+++ b/dumpling-overview.md
@@ -404,7 +404,7 @@ SET GLOBAL tidb_gc_life_time = '10m';
 | `--csv-separator`            | Separator of each value in CSV files. It is not recommended to use the default ','. It is recommended to use '\|+\|' or other uncommon character combinations| ','                                                                                                                                                                                                                                                                                               | ','                                        |
 | `--csv-null-value`           | Representation of null values in CSV files                                                                                                                                                                                                                                                                                         | "\\N"                                      |
 | `--csv-line-terminator`      | The terminator at the end of a line for CSV files. When exporting data to a CSV file, you can specify the desired terminator with this option. This option supports "\\r\\n" and "\\n". The default value is "\\r\\n", which is consistent with the earlier versions. Because quotes in bash have different escaping rules, if you want to specify LF (linefeed) as a terminator, you can use a syntax similar to `--csv-line-terminator $'\n'`. | "\\r\\n" |
-| `--csv-output-dialect`      | The value can be 'snowflake', 'redshift', or 'bigquery', indicating that the source data can be exported to CSV in the format required by the aforementioned databases. Such as binary data types will be converted to hexadecimal, but the '0x' prefix will be omitted. For example, '0x61' will be represented as '61'  | ""  |
+| `--csv-output-dialect`      | Indicates that the source data can be exported to a CSV file in a specific required format for the database. The option value can be `"snowflake"`, `"redshift"`, and `"bigquery"`. For example, if you set the option to `"snowflake"` or `"redshift"`, the binary data type will be converted to hexadecimal, but the `0x` prefix will be removed. For example, `0x61` will be represented as `61`. If you set the option to `"bigquery"`, the binary data type will be encoded using base64.  | ""  |
 | `--escape-backslash`         | Use backslash (`\`) to escape special characters in the export file                                                                                                                                                                                                                                                                | true                                       |
 | `--output-filename-template` | The filename templates represented in the format of [golang template](https://golang.org/pkg/text/template/#hdr-Arguments) <br/> Support the `{{.DB}}`, `{{.Table}}`, and `{{.Index}}` arguments <br/> The three arguments represent the database name, table name, and chunk ID of the data file                                  | `{{.DB}}.{{.Table}}.{{.Index}}`            |
 | `--status-addr`              | Dumpling's service address, including the address for Prometheus to pull metrics and pprof debugging                                                                                                                                                                                                                               | ":8281"                                    |