Releasesby GlyphModel Evaluations Python API for comparing model versions
Model evaluations is a new Python API for capturing how a model version performs against test data and visualizing results on the model page, outside of Modeling Objectives.
Also shipped Jul 7, 2026Palantir in July 2026 →
Deep Research in Snowflake CoWork (General availability)
Deep Research in Snowflake CoWork is now generally available. Deep Research is an investigation mode for complex, open-ended questions that require multi-step reasoning across your data.
REPLACE WHERE flows are now generally available
REPLACE WHERE flows for pipelines and standalone streaming tables are now generally available. Use them to replace only rows matching a predicate on each refresh, without reprocessing the full table history. See Batch processing with REPLACE WHERE flows .
Managed Iceberg materialized views are in Public Preview
You can now create a materialized view that is compatible with external Iceberg readers. This feature is in Public Preview . To enable it, reach out to your Databricks account team. See Create a materialized view compatible with external Iceberg readers .
[Launched] Generally Available: Network Security Perimeter support for Azure Event Hubs
Azure Event Hubs now supports Network Security Perimeter (NSP), enabling organizations to define a logical network isolation boundary for PaaS resources and control public network access through perimeter-based access rules.With Network Security Perimeter
[Launched] Generally Available: Confidential Computing support for Azure Event Hubs Dedicated
Azure Event Hubs Dedicated now supports Confidential Computing, enabling organizations to protect streaming data while it is being processed in memory. By leveraging hardware-based trusted execution environments (TEEs), customers can help safeguard sensit
Per-user budgets for cost centers in the billing UI
Enterprise admins can now create cost center user-level budgets directly in the billing UI where you manage cost centers and budgets. This feature is available for GitHub Enterprise Cloud. This gives you the same per-user AI credit budget controls that were previously only available through the REST API.
More from Foundry and AIPEvery Foundry and AIP release →
Claude Sonnet 5 available in AIP
Claude Sonnet 5 is available in AIP for eligible commercial and US government enrollments.
Pro-code agents authenticate against OSDK, Ontology MCP, and Palantir MCP
Foundry makes it easier to build, configure, and ship pro-code agents. Agents now authenticate against the Ontology SDK, Ontology MCP, and Palantir MCP with scoped permissions out of the box.
Automate supports Global Branching
Automate now supports Global Branching, so automations can be modified on a branch and tested end-to-end before merging into main.
Save AIP Analyst chats as analysis resources
AIP Analyst chats can now be saved as analysis resources in Compass to return to, share, and continue iterating, in standalone AIP Analyst and the Workshop widget.
Configurable per-user rate limits for AIP capacity
Enrollment administrators can now view and configure per-user rate limits for LLM usage in AIP.
Grok-4.5 from xAI available in AIP
Grok-4.5, xAI's newest model, is available in AIP with a 500k token context window.
GPT-5.6 series available in AIP
GPT-5.6 Sol, Terra, and Luna are available in AIP on US georestricted or non-georestricted enrollments with Direct OpenAI enabled, and on non-georestricted enrollments with Azure OpenAI enabled.
Marketplace installs and upgrades multiple products in one draft
Marketplace now supports installing and upgrading multiple products in the same installation draft, automatically fulfilling linked product inputs in the same job.