We’re introducing version 2 of the REST and SQL APIs for embedded Python and R execution in Autonomous AI Database, bringing important improvements including enhanced job tracking, new script filtering capabilities, and better timeout handling. As part of this update, version 1 endpoints are now deprecated and will be retired on April 30, 2026.
What’s New in Version 2
All Version 1 Functionality Preserved
Version 2 maintains full functional equivalence with all version 1 endpoints – you’ll simply update your endpoint paths from v1 to v2, as detailed in the migration section below.
Improved Asynchronous Job Tracking
Version 2 provides better visibility and control over your asynchronous jobs. You can now list your asynchronous jobs and delete any job whether running or completed. Version 1 allowed only deletion of running jobs. Additionally, we’ve enhanced asynchronous job handling for more efficient and reliable execution.
New Script Filtering Options
The new version also adds endpoints to filter your saved scripts by access type. Previously, you could list only scripts available to you. Now you can filter to view specific subsets:
- Scripts you own and saved to the script repository
- All scripts you have access to
- Global scripts accessible to all users
- Scripts you’ve granted access to others
- Scripts other users have shared with you
This makes it easier to locate specific scripts in your script repository.
Asynchronous Script Timeout Improvements
Version 2 handles job timeouts more gracefully. When the timeout limit is reached, the job status and result are now returned and persisted to the results table for later retrieval.
For SQL API Users
If you use the SQL API for embedded Python or R execution, the migration to v2 is transparent. The pyq*Eval functions remain the same, they will call the v2 rest endpoints internally.
What’s New in SQL API Version 2
In version 2, several new functions are added to improve your control over asynchronous jobs. You can now list your asynchronous jobs and delete running or completed jobs.
Checking for Deprecation Warnings
To see the deprecation warnings in v1 responses:
$ curl -i -X GET \s
--header "Authorization: Bearer ${token}" \
<OML Service URL>/oml/api/r-scripts/v1/script
Look for the Deprecation and Sunset headers in the output.
HTTP/1.1 200 OKDate: Fri, 13 Feb 2026 17:47:48 GMTContent-Type: application/json
Content-Length: 7079
Connection: keep-alive
Strict-Transport-Security: max-age=31536000;includeSubDomains
X-Content-Type-Options: nosniff
Warning: 299 - "This API version is deprecated. Use /v2/scripts instead."
Link: <OML Service URL>/oml/api/r-scripts/v2/scripts; rel="alternate"
Migrating to Version 2
To use the version 2 endpoints, update your REST endpoint path from v1 to v2 in the REST invocation. In this example, we use cURL to send a GET request to the scripts endpoint to list the scripts available to all users in the R script repository.
$ curl -s -X GET \
--header "Authorization: Bearer ${token}" \
/oml/api/r-scripts/v2/scripts/global | jq
{
"result": [
{
"name": "build.glm.mod",
"script": "function (df) \n{\n glm.mod <- ore.odmGLM(Petal.Length ~ ., IRIS_R)\n summary(glm.mod)\n}"
}…
What You Need to Know
Deprecation Timeline:
- Now through April 30, 2026: Version 1 endpoints remain functional but will return deprecation warnings in response headers.
- After April 30, 2026: Version 1 endpoints will not be available. All REST API calls for embedded Python and R execution must use version 2 endpoints.
We encourage you to begin migrating to version 2 during this deprecation period to ensure uninterrupted service.
How to Migrate
- Review the updated documentation for detailed API specifications
- Update your endpoint URLs from v1 to v2
- Test your integration in your development environment
- Complete your migration before April 30, 2026
For More Information
For complete API specifications and examples, refer to:
OML4Py REST API documentation for embedded execution
OML4R REST API documentation for embedded execution
OML4Py SQL API for embedded execution
OML4R SQL API for embedded execution
OML4Py Documentation
OML4R Documentation