Billing & Pricing
RAG Engine uses simple storage-based pricing. You pay for the documents you store — parsing, embedding, indexing, and search queries are all included.
Pricing Overview
RAG Storage
per GB per month
Included with your AppXen subscription. Billed as the rag_storage_gb dimension on your MCP Gateway Pro subscription.
What's Included
The per-GB price covers the full RAG pipeline — there are no additional charges for processing or queries:
- Document parsing — PDF, DOCX, HTML, Markdown, CSV, JSON, and more
- Chunking — intelligent text splitting with configurable overlap
- Embedding — vector embeddings via Amazon Bedrock Titan
- Vector indexing — pgvector storage and HNSW index maintenance
- Semantic search queries — unlimited, no per-query charge
How Storage Is Measured
Total document size
Storage is measured as the total size of your uploaded documents. This is the original file size, not the size of chunks or embeddings.
- Chunks and embeddings generated from your documents don't count extra
- Deleting a source removes its storage from your usage
Metered hourly as a gauge
Each hour, the system records your total storage. This is a gauge metric — it measures the current state, not cumulative usage. If you delete documents mid-month, your bill reflects the reduced storage from that point forward.
Cost Example
Typical knowledge base
A team uploads 500 MB of internal documentation (PDFs, Markdown, DOCX):
This includes unlimited semantic search queries against that knowledge base. Parsing and embedding are a one-time operation on upload — no recurring processing cost.
Subscription Channels
RAG Engine storage is billed through your AppXen subscription. Pricing is the same for both channels.
Stripe (Direct)
Subscribe at console.appxen.ai/subscribe with a credit card. Storage usage is metered hourly and billed monthly through Stripe.
- View storage usage in Settings → Billing & Usage
- Manage payment method and invoices via Manage Billing
- Cancel or resubscribe anytime from the console
AWS Marketplace (Coming soon)
Subscribe through the AWS Marketplace listing. Storage usage is metered hourly and appears on your AWS bill.
- Consolidated billing through your AWS account
- Manage subscription in the AWS Marketplace Console
- Use existing AWS payment methods and purchasing workflows
- Eligible for AWS Enterprise Discount Programs (EDPs)
Viewing Your Usage
Stripe customers
- Go to Settings in the console
- The Billing & Usage section shows your RAG storage alongside other dimensions
- The Knowledge Base page shows per-source document counts and sizes
AWS Marketplace customers (Coming soon)
Storage metering appears on your AWS bill:
- Go to AWS Billing Console
- Navigate to Bills → AWS Marketplace
- Find the
rag_storage_gbdimension under "AppXen for MCP"
Tips for Managing Costs
Delete unused sources
Remove documents you no longer need from the Knowledge Base. Storage drops immediately and your next billing hour reflects the change.
Prefer text over large PDFs
A 10 MB PDF with images takes 10 MB of storage. The same content as Markdown might be 200 KB. Use text formats when possible.
Use chunking wisely
Chunking settings affect search quality, not storage cost. Optimize chunk size for your content type rather than for cost.
Monitor in the console
Check the Knowledge Base page to see which sources use the most storage and identify candidates for cleanup.
Frequently Asked Questions
Do search queries cost anything?
No. Semantic search queries are unlimited and included in the per-GB storage price. There is no per-query charge.
Are there upload size limits?
Individual files can be up to 50 MB. There is no limit on total storage — you're billed for what you use.
What counts toward storage?
Only the original uploaded document size. Chunks, embeddings, and indexes generated by the RAG pipeline do not count toward your billed storage.
Does the file format affect cost?
Only indirectly. A 5 MB PDF costs the same as a 5 MB DOCX ($0.005/month). But the same content as plain text is typically much smaller, so text formats are more cost-efficient.