Databricks releases, every cloud
- Anthropic Claude Sonnet 4.6 now available as a Databricks-hosted model
Accessible via Foundation Model APIs on a pay-per-token basis.
- Serverless compute now available for compliance standards
Adds HITRUST, PCI-DSS, TISAX, and UK Cyber Essentials Plus support.
- Anthropic Claude Opus 4.6 now available as a Databricks-hosted model
Accessible via Foundation Model APIs, supports pay-per-token and query reasoning and vision models.
- New GCP regions support serverless compute
Available in me-central1 and southamerica-east1 regions.
- OpenAI GPT-5.2 Codex now available as a Databricks-hosted model
Excels at code generation and refactoring tasks. Accessible via Foundation Model APIs pay-per-token.
- OpenAI GPT-5.2 Codex now available as a Databricks-hosted model
Excels at code generation and debugging tasks, accessed via Foundation Model APIs pay-per-token.
- Run the change, do not just read it. Hands-on labs in your own Databricks workspace, graded when you submit. Browse labs
- Knowledge Assistant is now available in more regions
Available in us-east-2, ca-central-1, eu-central-1, eu-west-1, ap-southeast-1, and ap-southeast-2. Requires cross-geo processing in ap-southeast-1 and ap-southeast-2.
- You can now use custom base environments for Python, Python Wheels, and notebook tasks in serverless jobs
Defined with YAML files, they support Python, Python wheel, and notebook tasks.
- OpenAI GPT-5.1 Codex Max and Codex Mini now available as Databricks-hosted models
Access via Foundation Model APIs pay-per-token, subject to OpenAI's Acceptable Use Policy.
- OpenAI GPT-5.1 Codex Max and Codex Mini now available as Databricks-hosted models
Supports code generation and refactoring via Foundation Model APIs. Requires pay-per-token access.
- Anthropic Claude Haiku 4.5 now available as a Databricks-hosted model
Accessible via Foundation Model APIs on a pay-per-token basis.
- Gemini 3 Flash now available as a Databricks-hosted model
Offers speed and scale for video analysis and data extraction. Requires no setup.
- Serverless compute is now available in Azure China North 3
Available for notebooks, jobs, pipelines, and SQL warehouses.
- Serverless workspaces are now available in new GCP regions
Available in most GCP regions, see documentation for list.
- OpenAI GPT-5.2 now available as a Databricks-hosted model
Accessible via Foundation Model APIs on a pay-per-token basis.
- Anthropic Claude Opus 4.5 now available as a Databricks-hosted model
Accessible via Foundation Model APIs and AI Functions, supports pay-per-token and batch inference workloads.
- Serverless compute is now available in new regions
Added to Australia Southeast, Canada Central, Mexico Central, Norway East, and South Africa North regions.
- Google Gemini 3 Pro Preview now available as a Databricks-hosted model
Accessible via Foundation Model APIs and AI Functions, supporting pay-per-token and batch inference workloads.
- Get this table by email. One Monday mail covering the week, filtered to the clouds and products you run. Weekly digest
- Google Gemini 3 Pro Preview now available as a Databricks-hosted model
Accessible via Foundation Model APIs and AI Functions for pay-per-token and batch inference workloads.
- OpenAI GPT-5.1 now available as a Databricks-hosted model
Accessible via Foundation Model APIs and AI Functions for pay-per-token and batch inference workloads.
- New MCP server tab
Shows managed and external servers, and lists available servers from Databricks Marketplace.
- Serverless notebook tasks can now use jobs environment
Replaces inherited serverless environment with configurable job environment.
- Function calling and structured output now available using the Alibaba Cloud Qwen3-Next Instruct model (Beta)
Supports Qwen3-Next 80B A3B Instruct model.
- Supervisor Agent now supports Unity Catalog functions and external MCP servers
Delegates tasks to Unity Catalog functions and external MCP servers.
- Gemini 2.5 Pro and Flash models now supported on pay-per-token endpoints
Supports pay-per-token endpoints and Function calling.
- Serverless compute has been updated to version 17.3
Updated from version 17.2, requires no configuration changes.
- Improved autoscaling behavior for Mosaic AI Model Serving
Ignores brief traffic surges, scaling on sustained load increases. Reduces unnecessary concurrency scaling.
- SQL MCP server now available (Beta)
Executes SQL queries against Unity Catalog tables using SQL warehouses. Available at https://<workspace-hostname>/api/2.0/mcp/sql.
- The billable usage table now records the performance mode of serverless jobs and pipelines
Records performance mode in product_features.performance_target column with values PERFORMANCE_OPTIMIZED, STANDARD, or null.
- Mosaic AI Model Serving now supports OpenAI GPT 5 models (preview)
Supports GPT-5, GPT-5 mini, and GPT-5 nano models for batch inference with AI Functions.
- Anthropic Claude Sonnet 4.5 now available as a Databricks-hosted model
Accessible via Foundation Model APIs on a pay-per-token basis.
- Prompt caching is now supported for Claude models
Caches thinking messages, images, and tool use via cache_control parameter.
- Mosaic AI Agent Framework supports automatic authentication passthrough for Lakebase resources
Requires MLFlow 3.3.2 or above.
- Token-based rate limits now available on AI Gateway
Configures rate limits on model serving endpoints, replacing legacy governance settings, and requires configuration updates.
- OpenAI GPT OSS models now support function calling
Supports function and tool calling on Databricks-hosted models.
- OpenAI GPT OSS models now supported on provisioned throughput
Supports GPT OSS 120B and 20B models on provisioned throughput.
- Schedule jobs for Serverless GPU compute workloads
Creates and schedules Serverless GPU jobs via notebook scheduling.
- Use OpenAI GPT OSS models for batch inference and structured outputs
Optimized for AI Functions with batch inference and structured outputs. Supports ai_query().
- OpenAI GPT OSS models now available on Mosaic AI Model Serving
Supports GPT OSS 120B and 20B as hosted foundation models, accessible via pay-per-token APIs.
Headlines, dates and product areas are Databricks' own, and every item links to the note it came from. The one-line summaries are ours.