Skip to main content
Announcements

Introducing new Claude Opus 4 and Sonnet 4 models on Databricks

Build Smarter AI Agents on Databricks

by Ahmed Bilal, Hanlin Tang, Maddie Dawson and Ankit Mathur

Archived. This article has not been updated since the publish date above. The dynamic nature of information means that previously accurate content can become outdated or even obsolete over time. Readers are advised to exercise due diligence and cross-check any information found in this blog post before making decisions or adopting any practices based on said information.

  • Claude Opus 4 and Claude Sonnet 4 are now natively available in Databricks for customers across AWS, Azure, and GCP—offering secure, governed access to Anthropic’s most advanced model for reasoning, planning, and agentic tasks.
  • Build smarter, domain-specific AI agents by combining Claude’s hybrid reasoning with Databricks's tool integrations, unified API, and evaluation framework.
  • Run Claude directly in SQL or across Databricks tools like Notebooks and Workflows to power intelligent, context-aware applications at scale—no infrastructure required.

Reason over your data. Automate complex workflows. Scale with confidence — all in Databricks.

Two months after launching our partnership with Anthropic, we're thrilled to announce that Claude Opus 4 and Sonnet 4 are now natively available to Databricks customers on  AWS, Azure, and GCP. Opus 4 offers industry-leading reasoning and agentic capabilities, while Sonnet 4 provides an optimal balance of cost and performance for enterprise-scale deployments.

With full integration into the Databricks Data + AI Platform, enterprise teams can now securely build, evaluate, and govern Claude-powered AI agents using their own data — with centralized governance and observability.  This advances our mission to enable enterprises to build intelligent agents that can reason over their data.

What Claude 4 Unlocks for Enterprise AI

Claude Opus 4 — Intelligence for High-Stakes, Multi-Step Workflows

Claude Opus 4 is Anthropic’s most powerful model yet and a frontier model for coding, writing, and reasoning. It delivers sustained performance on long-running tasks that require focused effort and thousands of steps, with the ability to work continuously for several hours. Opus 4 combines frontier intelligence, improved memory capabilities, and tool use to enable autonomous agents that reason deeply, act across systems, and retain knowledge across sessions.  It is ideal for long-horizon work like engineering, legal review, or research synthesis.

Key Use Cases:

  • Advanced Coding: Plan and execute complex dev tasks like refactoring legacy codebases or building full-stack apps from spec.
  • Agentic Workflows: Power agents that manage multi-step business processes — like marketing campaigns or legal workflows.
  • Cross-Source Research: Analyze data from patents, filings, or reports to surface trends or conduct due diligence.
  • Virtual Collaborators: Build systems that retain memory across sessions, summarize prior work, and support sustained, multi-turn reasoning.

Content Creation: Generate long-form technical docs, natural marketing copy, and creative content with human-level fluency.

Claude Sonnet 4 — Fast, Scalable Intelligence for Everyday Automation

Claude Sonnet 4 brings impressive reasoning to high-throughput use cases — optimized for speed, efficiency, and production-scale deployments. It is ideal for real-time and operational automation.

Key Use Cases:

  • AI Assistants: Build real-time customer support and internal workflow agents that deliver accurate, context-aware responses.
  • Everyday Coding: Handle reviews, bug fixes, and API integrations with fast iteration and immediate feedback.
  • Business Intelligence & Analysis: Rapidly summarize dashboards, competitive data, and market signals.
  • Content at Scale: Create and analyze large volumes of enterprise content — from marketing assets to feedback reports.

Why Use Claude in Databricks?

Claude 4 models are now natively available in Databricks, allowing enterprises to securely build and scale AI systems over their private data — with no infrastructure to manage and governance built in.

  • Instant Access, No Setup: Claude Opus 4 and Sonnet 4 are ready to use via Model Serving — no deployment or configuration needed.
  • Unified and Simple Interfaces: Call Claude 4 through a single API or use it directly in SQL, Notebooks, Workflows, and DLT pipelines with AI Functions — no integration work required.
  • Enterprise-Grade Governance: All usage is automatically governed through the Agent Bricks AI Gateway — with built-in logging, safety guardrails, and PII detection. All request metadata and lineage is captured in Unity Catalog for unified access control and visibility.
  • Smarter Agents on Private Data: Build intelligent agents that reason over your data using tool calling, vector/metadata search, and orchestration graphs, all fully integrated with the Databricks platform.
  • Built-in Evaluation & Monitoring: Use Databricks’s evaluation framework to monitor and optimize model quality, latency, and cost in real workloads.

Read how Block uses Claude in Databricks to scale AI across teams — and how Claude can do the same for you.

 

Use Claude 4 in Python and SQL

Tool-Integrated Agents in Python

Batch Inference in SQL

What You Can Do Next

Claude 4 models will be rolling out soon to supported Databricks workspaces over the next few days.

Get the latest posts in your inbox

Subscribe to our blog and get the latest posts delivered to your inbox.

View all blogs