KNIME (Konstanz Information Miner) is an open-source data analytics, reporting, and integration platform that can handle various data sizes, from small to large datasets. The number of records KNIME can handle depends on several factors, including:

1. Hardware resources: The amount of available memory (RAM) and processing power (CPU) on your machine will significantly impact the number of records KNIME can handle efficiently. A more powerful machine can handle larger datasets.

2. Workflow design: The efficiency of your KNIME workflow can also influence the performance when handling large datasets. Optimizing the workflow and using appropriate nodes for specific tasks can improve the handling of larger datasets.

3. Data storage: How the data is stored and accessed, such as in-memory or on-disk, can impact the performance of KNIME when handling large datasets.

There is no specific limit to the number of records KNIME can handle, as it ultimately depends on the factors mentioned above. However, KNIME is capable of handling millions of records. To work efficiently with large datasets in KNIME, you can:

1. Increase the amount of available memory allocated to KNIME (by adjusting the knime.ini file).
2. Use streaming nodes, which can process data in smaller chunks, reducing memory requirements.
3. Optimize your workflow and use appropriate nodes for specific tasks.

It is important to note that for very large datasets (e.g., billions of records), you might need to consider alternative solutions, such as big data processing frameworks like Apache Spark or Hadoop, which are designed to handle and process massive datasets.