
Amazon Bedrock Pricing Plans & Tiers
Managed service to build generative AI apps on AWS
Pricing last verified: March 16, 2026
Pricing Analysis
Amazon Bedrock's pricing documentation lists 'Free' tier plus 'Standard, Flex, Priority, Reserved' options with zero dollar amounts published. This is intentional obfuscation: Amazon Bedrock is a managed service sitting on top of AWS infrastructure, and pricing is entirely usage-based and customer-specific. The 'tiers' listed are not product tiers but rather capacity reservation options (Standard = on-demand, Reserved = discounted capacity, Priority = SLA guarantees). Teams cannot estimate deployment costs without AWS account setup and actual inference runs.
The absence of any published per-token or per-inference costs is a critical pricing gap for SaaS teams evaluating Bedrock as an LLM inference platform. Competitors like OpenAI ($0.0001-$0.0015 per token), Anthropic ($0.003-$0.08 per token), and Google Gemini ($0.005-$0.03 per token) publish costs; Amazon Bedrock requires AWS account access and price estimation through AWS calculators. This opacity creates a competitive disadvantage vs. transparent pricing competitors.
Amazon Bedrock's positioning is not as an LLM API but as an AWS managed service for customers already embedded in AWS infrastructure. Pricing is bundled with AWS compute, storage, and networking costs rather than itemized as a separate line item. This integration is strategic—Amazon is optimizing for AWS account consolidation and preventing customer evaluation outside of AWS console.
Strengths
- Integration with AWS services (Lambda, SageMaker, data pipelines) enables seamless inference within AWS workloads without multi-cloud complexity.
- Reserved capacity and Priority tiers enable capacity reservation and SLA guarantees for production deployments.
- No per-seat licensing or minimum commitments—pricing is purely usage-based and scales with actual inference volume.
Considerations
- Entirely opaque pricing (no published costs) prevents pre-sales cost estimation and forces AWS account setup before evaluation.
- Integration with AWS infrastructure assumes customers are already on AWS—multi-cloud or non-AWS teams face migration costs.
- Lack of published per-token costs creates competitive disadvantage vs. OpenAI, Anthropic, and Google's transparent pricing.
AWS-native enterprises building generative AI workloads who want inference infrastructure integrated into existing AWS architecture.
Amazon Bedrock's opaque pricing ($) reflects its positioning as an AWS managed service, not a standalone LLM API.
Third-Party Ratings
Best choice: Amazon Bedrock
Try Amazon Bedrock freePricing Plans (102)
Jamba 1.5 Large
Jamba 1.5 Mini
Jurassic-2 Mid
Jurassic-2 Ultra
Jamba-Instruct
Provider
Anthropic
Cohere models
Rerank 3.5
**You are charged for number of queries where a query can contain up to 100 document chunks. If the query contains more than 100 document chunks, it is counted as multiple queries. For example, if a request contains 350 documents, it will be treated as 4 queries. Please note that each document can only contain upto 500 tokens (inclusive of the query and document’s total tokens), and if the token length is higher than 512 tokens, it is broken down into multiple documents.
Cohere Command
Cohere Command - Light
Embed 3 English
Embed 3 Multilingual
DeepSeek models
DeepSeek v3.2
DeepSeek v3.1
Google models
Gemma 3 4B
Gemma 3 12B
Gemma 3 27B
Meta models
Llama 2 Chat (13B)
Llama 2 Chat (70B)
Llama 2 Pretrained (13B)
Llama 2 Pretrained (70B)
Llama 2 Pretrained and Chat (13B)
Llama 2 Pretrained (70B)
Minimax models
Minimax M2
Minimax M2.1
Mistral models
Devstral 2 135B
Magistral Small 1.2
Voxtral Mini 1.0
Voxtral Small 1.0
Ministral 3B 3.0
Ministral 8B 3.0
Ministral 14B 3.0
Mistral Large 3
Kimi models
Kimi K2 Thinking
Kimi K2.5
NVIDIA models
NVIDIA Nemotron Nano 2
NVIDIA Nemotron Nano 2 VL
NVIDIA Nemotron 3 Nano 30B A3B
NVIDIA models
OpenAI models
gpt-oss-20b
gpt-oss-120b
GPT OSS Safeguard 20B
GPT OSS Safeguard 120B
Qwen models
Qwen3 Coder 30B A3B
Qwen3 32B
Qwen3 235B A22B 2507
Qwen3 Next 80B A3B
Qwen3 VL 235B A22B
Qwen3 Coder Next
Stability AI Image Services
Stable Image Remove Background
Stable Image Erase Object
Stable Image Control Structure
Stable Image Control Sketch
Stable Image Style Guide
Stable Image Search and Replace
Stable Image Inpaint
Stable Image Search and Recolor
Stable Image Style Transfer
Stable Image Conservative Upscale
Stable Image Creative upscale
Stable Image Fast Upscale
Stable Image Outpaint
Writer models
Palmyra X4
Palmyra X5
Z AI models
GLM 4.7
GLM 4.7 Flash
Custom Model Unit version
Price per Custom Model Unit per min*
Monthly storage cost per Custom Model Unit
Guardrails filter*
Content filters for both standard tier and classic tier (text content)
Content filters (image content)
Denied topics for both standard tier and classic tier
Sensitive information filters
Sensitive information filters (regular expression)
Word filters
Contextual grounding checks
Automated Reasoning checks
Model
Model selected for evaluation
Price Point
Intelligent Prompt Routing
Price per 1,000 tokens
$0.030
Claude Instant Inference
Claude 2.1 Inference
Human Tasks
Total
How does Amazon Bedrock pricing compare?
See how Amazon Bedrock's 102 pricing plans stack up against similar AI & ML tools.
Frequently Asked Questions
How much does Amazon Bedrock cost?
Does Amazon Bedrock offer a free plan?
What pricing model does Amazon Bedrock use?
Does Amazon Bedrock offer enterprise or custom pricing?
Track Amazon Bedrock Pricing Changes
Get notified when pricing changes for this tool and others you follow.
Reviews
No reviews yet. Be the first to review this tool.
Sources
- Amazon Bedrock Official Pricing— Vendor pricing page
- Amazon Bedrock Reviews— Independent reviews on G2
- Amazon Bedrock Reviews— Independent reviews on TrustRadius
Are you the team behind Amazon Bedrock?
Claim your profile to add custom descriptions, featured badges, and direct demo links.