How to manage large dataset in Power Query

Its completely depend on the steps you apply, but if you use UI and the tasks took such amount, I recomend you consider the below points:
Minimize Data Load
Filter Early: Apply filters as early as possible in your query to reduce the number of rows processed in subsequent steps.
Select Relevant Columns: Use the "Choose Columns" option to remove unnecessary columns right after importing data.
2. Optimize Query Design
Disable Background Preview: Disable "Enable Background Data Preview" in Power Query options to prevent unnecessary refreshes while editing.
Use Table.Buffer: For sorted or grouped operations, wrap the data in Table.Buffer to avoid re-evaluating steps multiple times.
Avoid Repeated Steps: Combine transformations into fewer steps to reduce dependency chains.
3. Leverage Native Queries
Push Operations to Source: Use SQL queries or other source-side operations (if supported) to filter, group, or transform data before it reaches Power Query.
Direct Query: When connecting to databases, write native SQL queries to retrieve only the required data.
4. Reduce Complexity
Avoid Nested Functions: Minimize the use of nested or custom functions in steps, as they can be slower with large datasets.
Simplify Logic: Break down complex logic into smaller, more efficient operations.
5. Manage Data Types
Set Data Types Once: Apply data type changes in the final steps, as Power Query evaluates this operation each time it's performed.
Remove Duplicates Smartly: Use deduplication only on necessary columns instead of entire rows.
6. Handle Merging and Joining Carefully
Sort Before Merge: Sort tables before merging to improve performance.
Reduce Table Size for Joins: Pre-filter data to include only rows that are necessary for merging.
7. Aggregate Efficiently
Group By Wisely: Avoid grouping by too many columns, and aggregate only the necessary metrics.
8. Use Power Query Settings
Increase Data Cache Memory: Go to Query Options > Data Load and increase cache memory to improve performance for large datasets.
Enable Parallel Loading: Allow parallel loading of tables in the query options.
9. Test with Sample Data
Work with Smaller Data Samples: Use a filtered subset of data while designing queries, then apply the full dataset for final execution.
10. Review Performance Metrics
Use Query Diagnostics: Analyze performance bottlenecks using Power Query's built-in Query Diagnostics feature.
Monitor Data Refresh Times: Identify steps that take the most time and optimize them.