Integrations
Connect every cost and usage signal
Track and optimize your AI, cloud, on-prem and data investments in one place, with new integrations continuously added.
Cloud 4
AWSServices, accounts and regions, Bedrock by model, and RI and Savings Plan coverage
See integration
AzureBilling accounts, subscriptions, resource groups and services
See integration
GCPBilling accounts, projects, services and labels
See integration
OracleTenancy and account cost for compute, network and block storage
Infrastructure 5
KubernetesClusters, namespaces, workloads and pods, across CPU, memory, GPU and storage
See integrationNutanix
Clusters, hosts and VMs, priced from utilization
NVIDIAGPU hours and utilization by cluster and team
OpenShift
Clusters, namespaces and workloads across CPU, memory and GPU
VMwareClusters, hosts, VMs and datastores, priced from utilization
See integration
AI platforms 7
AnthropicToken usage and cost by account, model and operation
See integration
AWS BedrockModel spend by account and operation, with inference profile tags
See integration
Bedrock AgentCoreCost and usage next to the agents and models it runs
Google Gemini
Cost and usage next to your other model providers
LiteLLMCost by provider, model, team and API key, plus agent traces
See integration
OpenAIToken usage and cost by account, model and operation
See integration
Vertex AIVertex AI spend from your Google Cloud billing account
See integration
Agentic 12
Agno
Cost and usage next to the models and tools your agents call
CrewAI
Cost and usage next to the models and tools your agents call
LangChainAgent steps and the model calls beneath them, by agent, user and customer
See integration
LangfuseTraces and observations by agent, model, user and session
See integration
LangGraphCost and usage next to the models and tools your agents call
Mavvrik Agent SDKCost per agent, customer and use case, including metered non-model usage
See integrationMicrosoft Agent Framework
Cost and usage next to the models and tools your agents call
Microsoft Copilot StudioAgent cost and credits by environment, user and model
n8nWorkflow runs, nodes and LLM calls, by workflow and customer
See integration
OpenAI AgentsCost and usage next to the models and tools your agents call
Semantic Kernel
Cost and usage next to the models and tools your agents call
Vercel EveAgent turns, model calls and tool calls, by agent and session
See integration
AI tools 5
ChatGPTCredits and cost by user, model and product, including Codex
See integration
Claude Code and CoworkCost and usage by user, team, session and model
See integration
CursorSeat and usage cost by user and team
GitHub CopilotSeats, usage and overage by user, team and cost center
See integration
Microsoft 365 CopilotLicense cost and assignments by user and department
See integration
Data platforms 5
ConfluentEnvironments, products and SKUs
See integration
DatabricksWorkspaces, clusters and service types from Databricks system tables
See integration
DatadogProducts, with on-demand and committed usage
See integration
MongoDBProjects, clusters and SKUs, with credits applied
See integration
SnowflakeWarehouses, accounts and service types across your organization
See integration
Vector databases 6
ChromaDB
Cost and usage next to the agents and models that use it
Milvus
Cost and usage next to the agents and models that use it
pgvector
Cost and usage next to the agents and models that use it
Pinecone
Cost and usage next to the agents and models that use it
Qdrant
Cost and usage next to the agents and models that use it
Weaviate
Cost and usage next to the agents and models that use it
Work management 6
Mavvrik sends cost alerts and findings to these tools.
Azure DevOpsCost alerts and findings sent to Azure DevOps
JiraCost alerts and savings findings as Jira tickets
Microsoft TeamsCost alerts in the channel that owns the spend
PagerDutyCost alerts sent to PagerDuty
ServiceNowCost alerts and findings sent to ServiceNow
SlackCost alerts in the channel that owns the spend
No integrations match that search. Ask us about it →
Not listed?
Bring in any other source.
The Mavvrik SDK, API, OpenTelemetry and MCP bring in cost and usage the connectors don’t cover yet, so it sits next to everything else.
Security
Enterprise controls from day one.
SOC 2 Type 2, single sign-on, role-based access so each team sees its own spend, and audit trails on every change and export.
FAQ
Questions you might have.
Can I see all of these costs in one place?
Yes. Every integration lands in the same platform, so cloud, data center, AI and data platform costs sit side by side under the same owners, and findings go to the people who own them through Jira, ServiceNow, Slack or Teams.
How does Mavvrik connect to each source?
Each integration uses the access the provider offers: a billing export or role for clouds, a usage API key for model providers, an appliance or agent for VMware and Kubernetes, and OpenTelemetry or the Mavvrik SDK for agents. Each integration page walks through the steps.
What data does Mavvrik receive?
Cost and usage data, such as billing exports, usage APIs and telemetry. Each integration page lists what Mavvrik receives from that source.
Can Mavvrik track AI agents and coding tools, not just model APIs?
Yes. Agent integrations such as LangChain, Langfuse, n8n and the Mavvrik Agent SDK tie each model call to the agent, user and customer behind it. AI tool integrations such as Claude Code, GitHub Copilot and ChatGPT show cost and usage by user and team.
Does Mavvrik cover on-prem and private cloud, or only public cloud?
Both. The VMware and Kubernetes integrations bring in clusters, hosts, VMs and workloads from your own data centers, priced from utilization, so on-prem and GPU infrastructure sit next to your public cloud and AI spend.
Can cost alerts go to the tools my teams already use?
Yes. Mavvrik sends cost alerts and findings to Slack, Microsoft Teams, Jira, ServiceNow, PagerDuty and Azure DevOps, so they reach the team that owns the spend.
What if a tool we use isn’t listed?
The Mavvrik SDK, API, OpenTelemetry and MCP bring in cost and usage from sources without a connector, so that spend sits next to everything else. We’re adding new integrations all the time, so tell us what you use and we’ll show you the best way to bring it in.