Amazon Bedrock Guardrails

Amazon Bedrock Guardrails is a cloud service for applying checks to model interactions. Its listed capabilities include prompt injection defense, output guardrails, PII redaction and secrets detection. Controls include text and image content filters, denied-topic rules, sensitive-information filters, word filters and contextual grounding checks. Separate InvokeGuardrailChecks options cover content filters, prompt attacks and sensitive-information filters; the content-filter and prompt-attack checks are specified for text only. The service is web-based and paid, with no free plan. Listed monthly prices include $0.15 a month for text content filters on Standard and Classic tiers and $0.15 a month for denied topics. Sensitive-information filters and contextual grounding checks each cost $0.10 a month. Image content filters, regular-expression sensitive-information filters and word filters are listed as free. InvokeGuardrailChecks content filters cost $0.07 a month, prompt-attack checks cost $0.08 a month, and sensitive-information checks cost $0.10 a month.

Amazon Bedrock Guardrails offers several types of checks, with individual listed prices for controls. It has no free plan, and some checks are specified for text only.

Amazon Bedrock Guardrails suits users seeking cloud-based checks for model interactions, including prompt attacks, sensitive information and output controls.

Amazon Bedrock Guardrails is a cloud service for applying safety and information checks to generative AI interactions. It is best suited to teams that need configurable safeguards for prompts and model outputs. Its broad set of controls is useful, though the per-control charges call for choosing protections to match the job.

Overview

The service combines content filters, denied topics, sensitive-information filters, word filters, and contextual grounding checks. It also supports prompt-injection defense, PII redaction, secrets detection, and output guardrails. That range can cover several safety needs in one cloud service, but the charges are divided among controls, so the cost depends on which checks a workload uses.

Key features

Content and topic controls

Text content filters and denied topics each cost $0.15 a month on Standard and Classic tiers. Image content filters cost nothing. This split lets teams choose protections by content type, while the text and topic checks carry separate charges. Word filters cost nothing, offering another control without a listed charge.

Sensitive information and grounding

Sensitive information filters cost $0.10 a month and support PII redaction and secrets detection. The regular-expression option costs nothing, which may suit teams whose needs can be handled with that filter. Contextual grounding checks also cost $0.10 a month, a separate charge for teams that need this kind of check.

Prompt attacks and output safeguards

Guardrails supports prompt-injection defense and output guardrails, extending its coverage to incoming prompts and generated responses. The InvokeGuardrailChecks prompt attack check costs $0.08 a month and is specified for text content only. The InvokeGuardrailChecks content-filter check costs $0.07 a month and is also text-only; its sensitive-information check costs $0.10 a month.

Pricing

Amazon Bedrock Guardrails is paid, with no free plan. Its listed controls are individually priced: text content filters and denied topics are $0.15 a month each on Standard and Classic tiers; image content filters are free; sensitive information filters and contextual grounding checks are $0.10 a month each; and word filters and regular-expression sensitive-information filters are free. The InvokeGuardrailChecks options cost $0.07 a month for content filters, $0.08 a month for prompt attack checks, and $0.10 a month for sensitive-information filters. The first two are text content only.

This structure favors teams that can select a defined set of checks. A workload using several paid controls incurs separate charges, while a team seeking a free plan cannot use Guardrails on that basis. There is no single plan price to compare against an all-in alternative.

Platforms

Guardrails is available on web and uses a cloud deployment model. That makes it a fit for teams choosing a hosted service rather than a self-hosted option.

Who it's for

Choose Guardrails when a cloud-based application needs a combination of prompt-injection defense, information filtering, and output safeguards, and when selecting and paying for individual checks suits the way the application is built. It is a weaker fit for readers who require a free plan or want a self-hosted deployment.

Pros and cons

• Pros: It brings content, topic, sensitive-information, grounding, and output controls together, with support for prompt-injection defense, PII redaction, and secrets detection.
• Pros: Several controls cost nothing or free, including image filters, word filters, and regular-expression sensitive-information filters.
• Cons: There is no free plan, and text content filters and denied topics each carry a $0.15 a month charge on Standard and Classic tiers.
• Cons: The InvokeGuardrailChecks content-filter and prompt-attack options are text-only, limiting their stated scope to text content.

Alternatives

Readers who want a free or open-source option can compare Sanctum Runtime, which is freemium and described as open-core with an MIT-licensed repository, or NVIDIA NeMo Guardrails, whose open-source Python package is free, though external models and services may need separate setup. Openlayer Guardrails has a free Basic plan for one member, five projects, one inference pipeline per project, 20,000 inference logs per month, 20 tests per project, one workspace, and three months of data retention; choose it when those caps fit.

PromptGuard offers a free tier with 20,000 scans a month, one API key, one project, and 24-hour log retention, plus a $19 a month Team plan. GLACIS offers a free account with no card required and setup described as taking about a minute. LLM Guard is a free, open-source solution, though its repository is archived and no longer actively maintained. Prompt Inspector has a free tier with 1,000 detections per month, REST API, SDK, and MCP access. CyberSecEval is a free open-source benchmark suite.

For broader comparisons, see the LLM Security Tools and AI Guardrail Software lists.

Verdict

Amazon Bedrock Guardrails is the right choice for teams that need a cloud service covering multiple AI safety controls and want to select them individually. Its prompt-injection defense, information controls, and output safeguards make a strong set; look elsewhere if a free plan or self-hosted deployment is essential.