ComputeCalcComputeCalc
AWS · Azure · GCP · Snowflake · Databricks

AI Pipeline
Cost Estimator

Model your end-to-end document AI spend across five clouds. Real pricing, transparent math, every number cited.

⚡ Live calculations
🔗 Cited sources
📊 5-cloud compare
Sign in to save

All cost figures are estimates based on publicly listed pricing as of April 2026. Hover any source chip for the official citation, or jump to the full Sources list ↓

Configuration

📁 Document Volume
6,000
200
500
300
🔍 OCR Strategy
🤖 LLM Provider

Default. Strong reasoning, vision, batch API.

🧠 LLM Model
🔢 Embedding Model
🗃️ Vector Store (6-month)
🏗️ Architecture View
Total Input Tokens
600.0M
tokens to process
Total Pages
1200K
6,000 × 200
Cost / Document
$0.82
all-in
Grand Total
$4.9K
incl. 6mo storage

Cost Breakdown by Component

Cost Estimate by Platform (same config)

Cost Breakdown
OCR Processing$4.8K
1200K pages
LLM Inference$34.50
Claude Haiku 4.5 · Batch · Cached
Embeddings$6.00
Amazon Titan Embed v2
Vector Storage (6mo)$105.47
585.9 GB · S3 Vectors
One-Time Processing$4.8K
+ Storage (6mo)$105.47
Total Estimate$4.9K
Cost per Document
$0.82
💚 Batch API saves $34.50
💡 Prompt Cache saves $270.00
🔷 S3 Vectors vs OpenSearch saves $738.28

Cost Distribution

Sources & References

Every cost figure in this estimator is derived from official, publicly listed pricing pages. Hover the source chip next to any option for the specific citation, or browse the full list below.

Anthropic Pricing
Haiku 4.5 input/output token rates
www.anthropic.com
AWS Bedrock Pricing
Same rates on Bedrock
aws.amazon.com
Vertex AI Pricing
text-embedding-005 token rate
cloud.google.com
AWS S3 Pricing
S3 Vectors storage — $0.03/GB/mo
aws.amazon.com
S3 Vectors Announcement
aws.amazon.com
Vertex AI Search Pricing
Index storage — $0.10/GB/mo
cloud.google.com
Pinecone Pricing
Serverless storage — $0.096/GB/mo
www.pinecone.io
OpenSearch Pricing
Serverless OCU minimums
aws.amazon.com
AWS Textract Pricing
AnalyzeDocument Forms — $0.05/pg, base $0.0015/pg
aws.amazon.com
Document AI Pricing
Form Parser tier
cloud.google.com
Anthropic Batches
Message Batches — 50% discount
docs.anthropic.com
AWS Bedrock Batch
docs.aws.amazon.com
Anthropic Prompt Caching
Cache reads at 10% of base input price
docs.anthropic.com
Azure AI Foundry Pricing
azure.microsoft.com
Doc Intelligence Pricing
azure.microsoft.com
Cortex Functions Pricing
Credit consumption (varies by region)
www.snowflake.com
Databricks Pricing
DBU rates for Model Serving, Vector Search, Mosaic AI
www.databricks.com
Mosaic AI Model Serving
Pay-per-token foundation model APIs
docs.databricks.com
Mosaic AI Vector Search
Storage & compute pricing
www.databricks.com

Pricing is volatile — confirm rates on the linked official pages before committing to a budget. Snapshot reflects April 2026.

Pricing reflects April 2026 rates · AWS Bedrock · Google Vertex AI · Azure MS Foundry · Snowflake Cortex
Claude Haiku 4.5 · Sonnet 4.5 · Opus 4.6 · Includes Batch API (50%) and Prompt Caching (90%) discounts