# New features
This release introduces Java 21 compilation support, upgrades the Studio platform to Eclipse 2026-03, and migrates Salesforce components to the Talend Component Kit framework.
## Shared features
| Feature | Description |
| --- | --- |
| Migration of Salesforce components and metadata to Talend Component Kit
framework | The Salesforce component family has been migrated to the Talend Component
Kit (TCK) framework. Existing Jobs remain backward compatible and can still be
migrated and run as expected.
Key changes:
- Salesforce metadata creation has been redesigned. Modules can no longer be created in bulk: datasets must now be created one at a time.
- In metadata, connection and dataset creation are now separated. The context menu action is now labeled Create Salesforce Dataset.
- In the Salesforce component settings, Connection type is renamed to Authentication type, with OAuth as the default value.
- The Salesforce URL and API version parameters are now in the Basic settings.
- The Use Proxy parameter is now labeled Use socks proxy.
- For tSalesforceOutput, tSalesforceBulkExec, and tSalesforceOutputBulkExec, the Main flow schema is no longer generated automatically. Click Guess Schema to generate it manually.
- When you switch login type, the default Connection URL value is not refreshed automatically. Click Default URL to update it.
- Parameter visibility follows the same logic as the previous Talend Component Kit framework where possible, with some parameter display behavior changes. For example, in tSalesforceOutputBulkExec, the Relationship mapping for upsert option is now labeled Relationship mapping: this parameter remains functional even when hidden for Insert and Update actions, so it is now displayed for both Insert and Update actions.
|
| Studio base Eclipse upgrade to R2026-03 | The Studio platform has been upgraded to Eclipse 2026-03 (4.39), bringing
the following improvements:
- Improved operating system support
- Fixes from the Eclipse platform
- Security updates
|
| Support for building Data Integration Jobs, Data Services and Routes with Java
21 | Talend Studio
now supports Java 21 to build Data integration Jobs, Data Services, and Routes.
- Java 21 is the recommended Java version for Talend 8.0.1-R2026-06 and later.
- Routine compilation level can be set to Java 8, Java 17, or Java 21.
- Data Integration Job, Data Service, Route, and Beans compilation level can be set to Java 17 or Java 21.
- The Routine compilation level must not be higher than the Job compilation level.
- Big Data Spark 4 Jobs are compiled with Java 17 compliance. Big Data Jobs remain compiled with Java 8 compliance but include add-opens parameters for Java 17 and Java 21 runtime compatibility in Talend modules.
Configure Java versions in File > Edit Project Properties > Build > Java Version. |
| Support for Debian 13 | You can now install and run Talend Studio on Debian 13. For more information,
see [Compatible Operating Systems](https://help.qlik.com/talend/en-US/installation-guide-windows/8.0/compatible-operating-systems). |
| New Talend Cloud
regions in Talend Studio | The following regions are now available in Talend Studio
when you select a region for remote license retrieval, remote connection setup,
and SSO login:
- United Kingdom on AWS
- United Arab Emirates on AWS
- India on AWS
- Ireland on AWS
|
| Support for more components for Job lineage in Qlik Cloud | Talend Studio
can now generate datasets and lineage to Qlik Cloud
for the following components:
- tMysqlOutputBulk
- tMysqlOutputBulkExec
- tMysqlBulkExec
- tPostgresqlOutputBulk
- tPostgreSQLOutputBulkExec
- tPostgresqlBulkExec
- tSAPHanaBulkExec
- tSingleStoreOutputBulk
- tSingleStoreOutputBulkExec
- tSingleStoreBulkExec
|
## Application Integration
| Feature | Description |
| --- | --- |
| Availability-note[Beta](https://help.talend.com/r/en-US/Cloud/labels-on-talend-help-center/availability-beta)
Beta content
New API-first components to build REST Data Services directly from API
contracts | The following components are now available to implement REST Data Services from
API contracts in a contract-first approach:
- tAPIContract to load the contract and generate operation flows.
- tAPIOperation and tAPIOperationResponse to implement operations.
|
| Enhancement of the memory support on cLangchainChat for Routes | The cLangchainChat component now provides enhanced memory support in Routes,
including isolated chat memory per session and persistence in relational or
NoSQL databases. You can configure conversation retention and optional session
TTL settings to control chat history lifecycle. |
## Big Data
| Feature | Description |
| --- | --- |
| Availability-note[Beta](https://help.talend.com/r/en-US/Cloud/labels-on-talend-help-center/availability-beta)
Beta content
Support for Databricks Serverless in Spark Batch Jobs
with serverless environment versions 4 and 5 | You can now run your Spark Batch Jobs on Databricks Serverless compute by
selecting Databricks Serverless from the Runtime
mode/environment dropdown list in the Spark
Configuration view. Serverless compute is fully managed by
Databricks and does not require any cluster configuration. It is available on AWS
and Azure cloud providers.
Requirements: Unity Catalog must be enabled, the
workspace must be in a region that supports serverless compute, and the PCI-DSS
compliance profile must not be enabled.
 |
| Support for CDP Public Cloud 7.3.2 with Spark 3.5.x | You can now run your Spark Jobs on a CDP Public Cloud Base 7.3.2 cluster
using Spark Universal with Spark 3.x in Cloudera Public
Cloud mode. You can configure it either in the Spark
Configuration view of your Spark Jobs or in the Hadoop
Cluster Connection metadata wizard.
 |
| Support for CDP Private Cloud Base 7.1.9 with Spark 3.5.x | You can now run your Spark Jobs on a CDP Private Cloud Base 7.1.9 cluster
using Spark Universal with Spark 3.x in Cloudera Private
Cloud mode. You can configure it either in the Spark
Configuration view of your Spark Jobs or in the Hadoop
Cluster Connection metadata wizard.
 |
| New Google Cloud Configuration in Metadata for Spark Batch Jobs | You can now create a Google Cloud Configuration from the Big Data
Configurations node in the Metadata folder
of the Repository tree view. The configuration stores your
Google Cloud Project ID and authentication credentials for reuse across Jobs. When
you drag and drop the configuration onto a Job, Talend Studio
creates a tGSConfiguration component in Spark Batch
Jobs. |
## Continuous Integration
| Feature | Description |
| --- | --- |
| Upgrade of Talend CI Builder to version 8.0.36 | Talend CI Builder is upgraded to version 8.0.36.
Use Talend CI Builder 8.0.36 in your CI commands or pipeline scripts from this monthly
version onwards until a new version of Talend CI Builder is released. |
## Data Integration
| Feature | Description |
| --- | --- |
| New tQVDInput component to read QVD files in Standard Jobs | A new tQVDInput component is available in Standard Jobs to read QVD files,
allowing you to reuse QVD data from Qlik Cloud and Qlik Sense directly in
Talend Studio workflows without converting files to intermediate formats first. |
| New option to set the page size for tNetSuiteInput (SOAP) in Standard Jobs | In the Advanced settings of tNetSuiteInput, you can now
set the number of records to retrieve per page when the component reads data
from the NetSuite SOAP service. |
## Data Mapper
| Feature | Description |
| --- | --- |
| Support for aggregation functions with conditions in a GROUP BY context | You can now use aggregation functions directly in conditions within a GROUP BY context. For example:
```
FROM order
GROUP BY custid AS cid
SELECT {
customer_id = cid,
order_status = if (count(orderno) == 1) 'One order' ELSE IF (count(orderno) == 2) 'Two orders' ELSE 'At least three orders'
}
ORDER BY cid
```
Aggregation functions in conditions are supported only within a GROUP BY context. They are not supported in WHERE clauses or other contexts outside GROUP BY. |
| New option to copy full standard DSQL script into clipboard in tDSQL4JSON | A new Copy Full Script to Clipboard option is
available in the tDSQL4JSON component allowing you to copy the DSQL script to the
clipboard. The copied script includes parameterized expressions but excludes
comments. |
## Data Quality
| Feature | Description |
| --- | --- |
| Support for Microsoft SQL Server for data mart | The data mart now supports Microsoft SQL Server 2019, 2022, and
2025. |
| Availability-note[Beta](https://help.talend.com/r/en-US/Cloud/labels-on-talend-help-center/availability-beta)
Beta content
New tSemanticMatch component for semantic similarity search | The new tSemanticMatch component enables semantic
similarity search against a vector database using an input column as the query
vector. The component connects to Milvus vector databases and retrieves the top
matching records based on similarity scores. |