Executable UDFs are generally available
Executable user-defined functions (UDFs) are now generally available on ClickHouse Cloud across AWS, GCP, and Azure, and are enabled for every organization. Since the public beta, UDFs have gained a native runtime for uploading precompiled, statically linked Rust, Go, C++, or Bun binaries; outbound network access for Python UDFs; a per-process memory limit...
Also shipped Oct 19, 2026ClickHouse in October 2026 →
Training Upcoming Serverless Training deprecation: DeepSeek V4 Flash (0731) On November 2, 2026 , DeepSeek V4...
Training Upcoming Serverless Training deprecation: DeepSeek V4 Flash (0731) On November 2, 2026 , DeepSeek V4 Flash (0731) will be removed from the shared Serverless Training pool. This change applies only to Serverless Training. If you train DeepSeek V4 Flash (0731) on Serverless Training, migrate to DeepSeek V4.1 Flash 1017 before November 2, 2026.
Cloud Workload Security is now available in Beta
Cloud Workload Security (CWS) is now available in Beta. Apply policy controls to cloud API requests before they reach your cloud provider. Define policies with request filters and Rego checks. For example, block requests that would make an R2 bucket public. Use workloads and policy sets to target your cloud environments.
Copilot model policy API for enterprise, orgs, and teams
You can now use the REST API to retrieve and update Copilot model policies for enterprises, enterprise teams, enterprise apps, and organizations. What’s new GET and PATCH endpoints are now generally available for model policies at four levels: Enterprise: /enterprises/{enterprise}/copilot/model-policies Enterprise teams...
GitHub Copilot models deprecated on October 19, 2026
As of today, October 19, 2026, we have deprecated the following models across most GitHub Copilot experiences (including Copilot Chat, inline edits, ask and agent modes, and code completions). These deprecations were announced on September 18, 2026 .
Send OpenTelemetry traces to Agent Observability, no SDK required
You can now send OpenTelemetry traces straight to Agent Observability, without the SDK. If your agents already emit OpenTelemetry generative AI spans, from your framework, an instrumentation library, or spans you write yourself, keep your existing route through Grafana Alloy or an OpenTelemetry Collector and opt in your AI spans with a single attribute.
Cortex Tools for Cortex AI Gateway (Preview)
We are pleased to announce the preview of Cortex Tools, a part of Cortex AI Gateway that connects the AI clients your organization uses to the business tools your work already lives in, using the Model Context Protocol (MCP). Supported clients include CoCo Desktop, CoCo CLI, Claude, and Cursor.
More from ClickHouse CloudEvery ClickHouse Cloud release →
Manage UDFs with the Cloud API and Terraform
You can now manage the full lifecycle of executable UDFs through 12 new Cloud API endpoints. The Terraform provider adds clickhouse_udf and clickhouse_udf_attachment resources in version 3.24.0 and later; both support pinning a UDF version per service. See the announcement and the UDF documentation.
UDF performance metrics in system.query_log
Starting in version 26.6, eight new ProfileEvents expose per-query UDF behaviour in system.query_log, including wall time, pool wait, CPU, memory, and bytes sent through the pipe.
Exactly-once delivery for Kafka ClickPipes
You can now choose exactly-once delivery when creating a Kafka ClickPipe instead of the default at-least-once mode. Recovery deduplicates replayed insert ranges while their deterministic tokens remain in the target table's deduplication window.
Tombstone handling for Kafka ClickPipes
Kafka ClickPipes can now handle tombstone messages: records with a key and a NULL payload. You can delete matching destination rows with lightweight deletes when exactly-once semantics is enabled, or mark tombstones as deleted rows.
Scheduled Chats in ClickHouse Agents (beta)
ClickHouse Agents can now run on a recurring schedule: hourly, daily, on weekdays, weekly, or using a custom cron schedule. Choose the time zone and days, and optionally organize each run in a Project. Manage schedules directly from the ClickHouse Agents sidebar.
Programmatic Tool Calling and in-run Skills for agents
Programmatic Tool Calling lets agents coordinate selected MCP tools through Code Interpreter with loops, branching, retries, and intermediate processing. To use it, enable Code Interpreter and mark tools as Programmatic in Agent Builder. Agents can also create reusable Skills while working and start using them in the same run.
Better management of long-running agent conversations
You can manually compact a conversation when its context gets crowded and see more clearly how much context tool results retain. Background tasks are easier to monitor and stop from the chat header, and supported results are delivered automatically. In the sidebar, you can filter and sort conversations, and pinned items now share one draggable section.
Exportable outputs, simpler file handling, and copyable shared conversations
Mermaid diagrams open as full artifacts and can be exported to SVG or PNG; other artifacts can be viewed full-screen, and code blocks can be downloaded as files. You can send an attachment without a prompt, and agents route files to the appropriate tools only when needed. You can also copy a shared conversation into your account and continue it as your own.