What is Amazon Bedrock Studio? What it was, and where it lives now (Amazon Bedrock in SageMaker Unified Studio)

What is Amazon Bedrock Studio? What it was, and where it lives now

If you are weighing the wider environment rather than just this builder, our guide to SageMaker Unified Studio vs Bedrock covers how the unified workspace relates to Bedrock itself. Here we focus on the Bedrock Studio question specifically: what happened to it, and what to use now.

The short answer

“Amazon Bedrock Studio” was the name of a standalone, low-code builder for Bedrock-based generative AI apps. AWS has since consolidated its data and AI tooling into one place, and that builder now lives inside SageMaker Unified Studio under the name Amazon Bedrock in SageMaker Unified Studio.

So if a tutorial, a conference talk, or an old bookmark points you at “Bedrock Studio”, read it as the same idea under a new home. The underlying Bedrock service - models, Knowledge Bases, Guardrails, Agents - is unchanged. What moved is the low-code front end you use to assemble them into an application.

This is the same pattern we have seen across the AI-platform space in 2026: tools getting renamed or absorbed into larger suites faster than the documentation and the search results can keep up. Google did it when Vertex AI became the Gemini Enterprise Agent Platform, and the confusion it creates is the same - the product is fine, the name moved.

What Amazon Bedrock Studio was

Bedrock Studio was introduced as a way to let people build generative AI applications on Amazon Bedrock without managing the plumbing underneath. The pitch was accessibility: a developer, an analyst, or a product person could prototype a RAG chatbot or an agent in a browser, share it with colleagues, and iterate - without standing up infrastructure for each piece.

The building blocks it exposed are the ones Bedrock itself provides:

• Knowledge Bases - connect your own data sources and use Retrieval Augmented Generation (RAG) so model answers are grounded in your content.
• Agents - build chat or task agents that can call tools and follow multi-step instructions.
• Guardrails - apply safety and privacy filters to inputs and outputs.
• Prompts and Flows - manage prompt templates and chain steps into a working application.
• Functions - extend an app with custom logic and integrations.

The value was never a new model or a new engine. It was the governed, low-code surface that let a team go from idea to a shareable prototype quickly, on top of Bedrock.

What changed - and why

AWS spent 2026 consolidating a scattered set of data and AI tools into a single environment, SageMaker Unified Studio, which became generally available during the year. Rather than keep a separate standalone Bedrock Studio alongside it, AWS brought the same low-code gen-AI app-building experience inside the unified studio.

The result is Amazon Bedrock in SageMaker Unified Studio: the same Knowledge Bases, Agents, Guardrails, Prompts, Flows, and Functions, now reached through one workspace that also holds SQL analytics, data processing, and model development, with a shared catalogue and shared governance across all of it.

The reasoning is the same one behind the whole unified-studio push: fewer disconnected consoles, one place to control who can access what, and a straight line from your data to a GenAI application without hopping between tools. For a team, that mostly removes the seams - the governance and the catalogue follow you from a SQL query to a Knowledge Base to a deployed agent.

What it does now

Inside SageMaker Unified Studio, the Bedrock experience keeps its original job: build generative AI applications through a simple, accessible interface, without managing the underlying services. In practice that means you can:

• Create Knowledge Bases from your own data for RAG, so answers are grounded in your content rather than the model’s general training.
• Build chat and task Agents, and attach Guardrails for safety and privacy.
• Use prompt engineering, Functions, and Flows to turn a model into an actual application.
• Collaborate across business and technical roles in one governed workspace, and share apps securely.

The difference from the standalone days is context, not capability: you are doing this inside the same environment where your data and your other AI work already live, under one set of access rules.

Source: the AWS Amazon Bedrock in SageMaker Unified Studio product page.

When to use it (and when to go to Bedrock directly)

The low-code builder earns its place when the goal is speed and collaboration rather than fine-grained control:

• Reach for Bedrock in SageMaker Unified Studio when you want to prototype or ship a RAG app or an agent quickly, let non-specialists participate, and keep everything under shared governance. It is the fastest path from data to a working GenAI app for most teams.
• Go to Amazon Bedrock directly (API/SDK) when you need programmatic control, want the model layer embedded in your own application or pipeline, or are building something the low-code surface does not express. Our runbook on handling Bedrock 503 capacity errors is the kind of production concern you manage at that level.
• Think about the whole environment, not just the builder, when the decision is really about where your team works day to day. That is the SageMaker Unified Studio vs Bedrock question, and it is worth answering deliberately rather than by default.

If you are designing a production generative AI layer on AWS - RAG, agents, guardrails, and the cost and reliability work underneath - that is exactly the kind of build we help with. See our cloud cost optimisation work for how we keep those workloads predictable, or get in touch below.

Frequently asked questions about Amazon Bedrock Studio

Building a generative AI app on Bedrock?

Book a free 30-minute call. We'll help you decide between the low-code studio and building on Bedrock directly, and design the RAG, agent, and cost-control layers so the result is production-ready.

Book a call Send us a message