Integrations · AI platforms
See Bedrock cost with all your other spend
Track Bedrock cost by account, model and operation, split by team or app through inference profile tags, in the same place as the rest of your cloud, data and AI spend.

Bedrock cost and usage
Bedrock on-demand model calls cannot be tagged
Bedrock cost lands in your AWS bill by account and model, with no way to tag on-demand calls. Without inference profiles, it is hard to say which team or app the spend belongs to.
- Bedrock spend mixed into the AWS billBedrock cost by account, model and operation in the GenAI cost dashboard
- No way to tag on-demand callsCost by application, team or environment through inference profile tags
- Agent cost unclearAgent-level Bedrock and AgentCore cost through the Mavvrik SDK
- Overruns found after the invoiceBudgets, alerts and cost allocation on GenAI spend
What you can do
See, split and budget Bedrock cost
Group by account, model, operation, usage type, location or inference profile tag
- support-assistant$14.1K
- doc-search$9.8K
- claims-review$7.3K
- Untagged$5.2K
- intake-agent$4.2K
- triage-agent$2.9K
See everything
Cost by model and operation
See Bedrock cost by account, model and operation in the GenAI cost dashboard.
Cost by application
Split spend by application, team or environment with the tags on your Application Inference Profiles. Tags show up within 24 hours of activation.
Agent-level cost
Instrument Bedrock and Bedrock AgentCore calls with the Mavvrik SDK to see cost by agent, session, user, customer and use case.
Control costs
Budgets and alerts
Set budgets and alerts on GenAI spend.
Cost allocation
Allocate Bedrock spend with the rest of your AWS cost.
How it works
Bedrock comes in with your AWS billing account
There is no separate credential. Once your AWS billing account is connected, Mavvrik finds accounts with AI services and connects them. Mavvrik holds a current SOC 2 Type 2 report and supports single sign-on and role-based access.
1 Connect
Connect your AWS billing account
Mavvrik finds accounts with AI services and connects them. Auto-connect is on by default.
2 Data flows in
Tag inference profiles for apps
Create Application Inference Profiles with tags, activate the tags in AWS Cost Allocation Tags, and invoke models with the profile ARN.
3 See your costs
Bedrock cost within 24 hours
The GenAI cost dashboard shows data within 24 hours of connecting. Inference profile tags show up within 24 hours of activation.
Total cost
Bedrock is just one part of your total AI cost
Some of the other costs you may want to track alongside Bedrock
CloudAWSSee AWS cost by service, account and team, including Bedrock.View integration →
AI platformsAnthropicSee Claude API cost by model and operation.View integration →
InfrastructureKubernetesSee cluster cost by namespace, node and label.View integration →
AgentsLangChainSee cost per agent and workflow, traced to each model call.View integration →
FAQ
Questions you might have
Do we need a separate Bedrock connection?
No. Bedrock spend comes from your connected AWS billing account.
How do we split Bedrock cost by team?
Bedrock does not allow tags on on-demand model calls. Create Application Inference Profiles with tags, activate the tags in AWS Cost Allocation Tags, and call models through the profile ARN. Mavvrik reads those tags.
Can we see cost per agent?
Yes, with the Mavvrik SDK. It instruments Bedrock and Bedrock AgentCore calls in Python and JavaScript or TypeScript.