cloudcost-cli

CloudCost AI provides a solution for understanding and managing cloud infrastructure and AI model usage costs at critical decision points, such as before merging code or making API calls. The platform is designed to integrate with existing developer workflows, surfacing cost estimates directly within GitHub pull requests and tracking large language model (LLM) spending across various providers.

For infrastructure changes managed through Terraform, CloudCost AI operates by reading Terraform plan JSON files generated in continuous integration pipelines. It uses Infracost to price each resource and then posts a detailed monthly cost delta as a comment on the relevant pull request. This breakdown includes resource-level insights and highlights the largest cost drivers, supporting AWS, Google Cloud, Azure, and Kubernetes manifests. The tool emphasizes visibility by ensuring cost information appears where code reviews occur, rather than requiring teams to consult separate dashboards.

In addition to infrastructure cost analysis, CloudCost AI includes an LLM gateway that meters API usage across providers such as OpenAI, Anthropic, Google, Gemini, Mistral, Meta Llama, AWS Bedrock, and Azure OpenAI. It issues virtual keys, enforces per-team budgets, and tracks token usage and costs by user, team, feature, or agent. The gateway is OpenAI-compatible and does not store prompt or completion data by default, focusing on privacy and compliance. Telemetry records token counts, model names, and cost metadata, enabling granular tracking of LLM spend without storing sensitive payloads.

CloudCost AI can be self-hosted, allowing organizations to run the pricing API within their own network and point the Infracost CLI at a local Cloud Pricing API. The platform is designed to fit into existing workflows with minimal setup, requiring only configuration of GitHub and OpenAI-compatible clients.