Amazon Bedrock

What it does

Amazon Bedrock is a managed AWS service for building generative AI applications and agents. It takes in text, image and other prompts and sends them to foundation models from multiple providers, returning generated text, images, video, speech, embeddings and agent actions. Developers can add their own data through knowledge bases, prompt engineering or fine-tuning, and attach guardrails that filter harmful content or check responses against defined policies.

People use it through the AWS console playgrounds, the Bedrock API, the AWS CLI or SDKs to build chat assistants, document summarization, copywriting, image generation and workflow automation. Distinguishing features include model choice across many providers, evaluation tools for comparing models, Amazon Bedrock AgentCore for building and deploying agents, and cost controls such as batch inference, prompt caching, model distillation and intelligent prompt routing. Monitoring and logging support governance and audit needs.

It suits developers and enterprise teams working in AWS who need security, compliance and monitoring for production AI. Access requires an AWS account and a request for model access. Pricing varies by model, provider and region, so review per-model terms and regional availability before committing.

What it can do

• Access to foundation models from multiple AI providers, with evaluation tools to compare models
• Knowledge bases and fine-tuning to customize models with your own data
• Bedrock Guardrails for content filtering and automated reasoning checks on responses
• Agent development through Amazon Bedrock AgentCore for building and deploying agents
• Cost optimization features including batch inference, prompt caching, model distillation and intelligent prompt routing
• Monitoring and logging for governance and audit requirements

Works with

• AWS Command Line Interface
• AWS SDK for Python (Boto3)
• Amazon SageMaker Unified Studio
• Amazon SageMaker AI notebooks

Example workflows

Customer support assistant (illustrative)

Input: customer questions and product documentation loaded into a knowledge base. Steps: select a model in the playground, attach a guardrail, test responses, then deploy through the API. Output: a chat endpoint that answers questions grounded in the supplied documents.

Document summarization (illustrative)

Input: long reports, research papers or technical documentation. Steps: send the document as a prompt to a chosen model, optionally using batch inference for large volumes. Output: concise summaries that extract key information.

Before you choose

• Pricing is per model, provider and region, with separate rates for on-demand, batch, flex, priority and reserved tiers; check the rate table for your specific model.
• Some models, such as certain Anthropic Claude Mythos versions, are gated and require approval from the provider's account team.
• Provisioned Throughput and custom model pricing require contacting an AWS account or sales team to request a quote.

Listed under

• Coding
• AI agents
• Developers
• Enterprise
• Foundation models
• Genai
• Model API

Get started

Start from the getting-started page; new AWS customers receive up to $200 in credits for trying AWS AI.

1. Sign in to the AWS console and set up the Amazon Bedrock IAM role.
2. Request access to the foundation models you intend to use from the Bedrock console.
3. Generate an API key or configure the AWS CLI or SDK to make your first model invocation.

New AWS customers receive up to $200 in AWS credits for trying AWS AI; this is a time-limited allowance, not a permanent free tier. Model prices are per million tokens, per image, per 1,000 queries or per hour depending on the product. Batch and Flex are listed at 50% of Standard tier pricing, and Priority is listed at a 75% premium. Verify exact rates in the pricing rate table for your model and region.

Source: aws.amazon.com/bedrock/pricing, checked 28 Oct 2026. Prices can change; check before you buy.

Sources for this overview

Product pages, documentation, and pricing checked 28 Oct 2026. Each source supports part of the overview; availability and prices can change.

• Amazon Bedrock – Build genAI applications and agents at production scale – AWS ↗aws.amazon.com
• Amazon Bedrock Pricing ↗aws.amazon.com
• Get started with Amazon Bedrock – Amazon Bedrock – AWS ↗aws.amazon.com

Company website

aws.amazon.com: product, pricing and policy pages

YC directory mirror

Batch and company name

Web search

Confirms the official site

Social accounts

Accounts the company links, and its posts on X

Human review

Checked by a person

knowledge_base

Last checked 28 Oct 2026. Every field keeps its source and date; unknown means we couldn’t confirm it.