Calculate your AI costs
Model your AWS Bedrock costs

Use this calculator for a ballpark estimate of your AWS Bedrock spend. Click on the workload(s) your application will use and enter the details for the expected volume and specs. As you enter your data, the summary will automatically update.

Note: Actual usage can vary based on input/output requirements, prompt optimization, model routing, batch processing, and various price-optimization techniques within AWS.
Chat Tokens/Month
Doc Tokens/Month
Image Tokens/Month
Total Tokens/Month
Provider Model Monthly Cost Annual Cost

Anthropic Models

Model Tokens per minute (TPM) Requests per minute
Claude 3 Haiku 2,000,000 (us-east-1, us-west-2)
300,000 (other regions)		 1,000 (us-east-1, us-west-2)
400 (other regions)
Claude 3 Opus 400,000 50
Claude 3 Sonnet 1,000,000 (us-east-1, us-west-2)
200,000 (other regions)		 500 (us-east-1, us-west-2)
100 (other regions)
Claude 3.5 Haiku 2,000,000 (most regions)
300,000 (us-west-1)		 1,000 (most regions)
400 (us-west-1)
Claude 3.5 Sonnet 2,000,000 (us-west-2)
400,000 (select regions)
200,000 (others)		 250 (us-west-2)
50 (select regions)
20 (others)
Claude 3.5 Sonnet V2 2,000,000 (us-west-2)
400,000 (other regions)		 250 (us-west-2)
50 (other regions)
Claude 3.7 Sonnet V1 1,000,000 (cross-region) 250 (cross-region)
Claude Instant 1,000,000 (us-east-1, us-west-2)
300,000 (other regions)		 1,000 (us-east-1, us-west-2)
400 (other regions)
Claude V2 500,000 (us-east-1, us-west-2)
200,000 (other regions)		 500 (us-east-1, us-west-2)
100 (other regions)

Amazon Models

Model Tokens per minute (TPM) Requests per minute
Nova Lite 2,000,000 1,000
Nova Micro 2,000,000 1,000
Nova Pro 400,000 100
Titan Text Express 300,000 400
Titan Text Lite 300,000 800
Titan Text Premier 300,000 100
Titan Text Embeddings V2 300,000 2,000
Titan Multimodal Embeddings G1 300,000 2,000
Titan Image Generator G1 2,000 60
Titan Image Generator G1 V2 2,000 60

Meta Models

Model Tokens per minute (TPM) Requests per minute
Llama 2 13B 300,000 800
Llama 2 70B 300,000 400
Llama 2 Chat 13B 300,000 800
Llama 2 Chat 70B 300,000 400
Llama 3 8B Instruct 300,000 800
Llama 3 70B Instruct 300,000 400
Llama 3.1 8B Instruct 300,000 800
Llama 3.1 70B Instruct 300,000 400
Llama 3.1 405B Instruct 400,000 200
Llama 3.2 1B Instruct 300,000 800
Llama 3.2 3B Instruct 300,000 800
Llama 3.2 11B Instruct 300,000 400
Llama 3.2 90B Instruct 300,000 400
Llama 3.3 70B Instruct 300,000 400
Chat Tokens/Month
Doc Tokens/Month
Image Tokens/Month
Total Tokens/Month
Provider Model Monthly Cost Annual Cost

Anthropic Models

Model Tokens per minute (TPM) Requests per minute
Claude 3 Haiku 2,000,000 (us-east-1, us-west-2)
300,000 (other regions)		 1,000 (us-east-1, us-west-2)
400 (other regions)
Claude 3 Opus 400,000 50
Claude 3 Sonnet 1,000,000 (us-east-1, us-west-2)
200,000 (other regions)		 500 (us-east-1, us-west-2)
100 (other regions)
Claude 3.5 Haiku 2,000,000 (most regions)
300,000 (us-west-1)		 1,000 (most regions)
400 (us-west-1)
Claude 3.5 Sonnet 2,000,000 (us-west-2)
400,000 (select regions)
200,000 (others)		 250 (us-west-2)
50 (select regions)
20 (others)
Claude 3.5 Sonnet V2 2,000,000 (us-west-2)
400,000 (other regions)		 250 (us-west-2)
50 (other regions)
Claude 3.7 Sonnet V1 1,000,000 (cross-region) 250 (cross-region)
Claude Instant 1,000,000 (us-east-1, us-west-2)
300,000 (other regions)		 1,000 (us-east-1, us-west-2)
400 (other regions)
Claude V2 500,000 (us-east-1, us-west-2)
200,000 (other regions)		 500 (us-east-1, us-west-2)
100 (other regions)

Amazon Models

Nova Lite 2,000,000 1,000
Nova Micro 2,000,000 1,000
Nova Pro 400,000 100
Titan Text Express 300,000 400
Titan Text Lite 300,000 800
Titan Text Premier 300,000 100
Titan Text Embeddings V2 300,000 2,000
Titan Multimodal Embeddings G1 300,000 2,000
Titan Image Generator G1 2,000 60
Titan Image Generator G1 V2 2,000 60

Meta Models

Llama 2 13B 300,000 800
Llama 2 70B 300,000 400
Llama 2 Chat 13B 300,000 800
Llama 2 Chat 70B 300,000 400
Llama 3 8B Instruct 300,000 800
Llama 3 70B Instruct 300,000 400
Llama 3.1 8B Instruct 300,000 800
Llama 3.1 70B Instruct 300,000 400
Llama 3.1 405B Instruct 400,000 200
Llama 3.2 1B Instruct 300,000 800
Llama 3.2 3B Instruct 300,000 800
Llama 3.2 11B Instruct 300,000 400
Llama 3.2 90B Instruct 300,000 400
Llama 3.3 70B Instruct 300,000 400

We help funded startups and high-growth teams design, build, and scale AI applications — leveraging AWS funding programs to lower your costs.