Releasesby GlyphPersistent memory for eve agents
eve agents can now retain context across sessions and use it in future conversations. Persistent memory is organized into slots. You can define named slots in files under agent/memory/ . Each slot specifies a provider, which stores and retrieves the memory, and a scope, which determines who or what shares it.
Also shipped Sep 9, 2026Vercel in September 2026 →
Configurable retention for system tables is in Beta
Account admins can now set an account-wide retention period from 30 to 3,650 days for supported system tables. See Configure system table data retention .
External secrets in Unity Catalog (Beta)
External secrets connect a Unity Catalog schema to AWS Secrets Manager. The secrets appear in Unity Catalog as read-only securable objects that you govern with Unity Catalog privileges. Databricks reads their values from AWS Secrets Manager on demand. See External secrets in Unity Catalog .
Celigo connector (Beta)
The Celigo connector is now available in Lakeflow Connect in Beta . Workspace admins can control access to this feature from the Previews page. The connector ingests audit log events, such as create, update, and delete operations on resources, from Celigo (integrator.io) into Databricks. Authentication uses a Celigo integrator.io API token.
Anysphere Organization connector (Beta)
The Anysphere Organization connector is now available in Lakeflow Connect in Beta . Workspace admins can control access to this feature from the Previews page. The connector uses the Anysphere (Cursor) Organization Admin API to ingest organization members, groups, and group members into Databricks as daily snapshots.
Anaplan connector (Beta)
The Anaplan connector is now available in Lakeflow Connect in Beta . Workspace admins can control access to this feature from the Previews page. The connector ingests audit trail events and user account records from your Anaplan tenant into Databricks. Authentication uses an Anaplan CA certificate and its paired private key. See Anaplan connector .
Openflow gen 1 deployment creation retired
You can no longer create new first generation (gen 1) Openflow deployments (BYOC or Snowflake). All new deployments are second generation (gen 2), created with CREATE OPENFLOW DEPLOYMENT or the Openflow UI. Existing gen 1 deployments continue to work unchanged, and you can still create new gen 1 runtimes on an existing gen 1 deployment.
More from AI SDK and AI GatewayEvery AI SDK and AI Gateway release →
DeepSeek V4.1 Flash now available on AI Gateway
DeepSeek V4.1 Flash is now available on AI Gateway with native image understanding. V4.1 Flash has vision support and accepts text and images in the same request, so you can ask questions about screenshots, read charts, and extract information from visual content.
GitHub Copilot is now available in the AI SDK harness layer
The AI SDK harness layer now supports GitHub Copilot through the official @ai-sdk/harness-github-copilot adapter. The harness layer lets your application run different coding agents through the same HarnessAgent interface, so you can switch agents without changing your application code.
Build with OpenAI Agents API on Vercel
You can now build and deploy long-running, tool-using agents with the OpenAI Agents API on Vercel. OpenAI manages the agent loop and session state, while Vercel hosts the application and connects each session to Vercel Sandbox for code execution and file access.
Tako Search is free on AI Gateway through September 30
Tako Search is free exclusively on AI Gateway through September 30. It lets AI models search Tako's curated data and the live web, filter web results by domain or publication date, and use the results to answer questions with current information, citations, and visualizations. After September 30, searches are billed at standard rates.
Vercel Sandbox is now available in all regions
Vercel Sandbox can now run in all 20 Vercel compute regions , up from four. Running sandboxes closer to the databases, storage, and other services they access reduces latency. Teams can also keep sandbox workloads in approved regions to support data residency and regional processing requirements. iad1 remains the default region.
Vercel Sandbox routing is now 18x faster globally
Requests to Vercel Sandbox public domains are now routed 18x faster. Domains created with the sandbox.domain() SDK call are now resolved from the nearest regional replica, instead of a single centralized store. Incoming requests reach the process running in the sandbox with less latency. Median domain lookup latency dropped from 62ms to 3.4ms (18x faster).
GPT Image 2.5 Flare and Sunburst now available on AI Gateway
GPT Image 2.5 Flare and GPT Image 2.5 Sunburst from OpenAI are now available on AI Gateway. Both models accept text prompts and reference images for generation and editing.
Vercel Sandbox now provides 64 GB of storage
Every Vercel Sandbox now includes 64 GB of storage, up from 32 GB. This includes sandboxes created from a Vercel Managed Image or custom image, as well as those configured with the deprecated runtime property.