1
Previously product @ AWS on container services, product lead for compute @ DigitalOcean.
I love infrastructure - please feel free to contact me at ben@vantage.sh
- - - -
I'm Co-Founder and CEO of Vantage posting on behalf of our team that shipped this. Vantage is a cloud cost management platform. We also maintain https://ec2instances.info/ after buying the site a few years back. We've been hearing rumblings from the community looking for something similar for LLMs...so we built it.
The site is free and open-source and provides a centralized view of LLM model specifications, pricing, benchmarks, and hosting information. We're shipping early and looking for feedback. Give it a peek and let us know what you think.
Our team recently launched the ability for you to be alerted about AWS pricing changes for existing EC2 instances as well as alerts for new instances being launched and what their associated pricing is.
We've had this on in the background for a few weeks and it's alerting for new instance types faster than AWS' blog posts hit. Also whenever anything changes for existing EC2 instances you can see as soon as it changes.
If you want to set an alert, you can see this here: https://newsletters.vantage.sh/
I’m on the team at www.vantage.sh and we’re releasing our remote MCP hosted on Cloudflare Workers that gives LLMs context about your structured cost and usage data on AWS, Azure, GCP, OpenAI, Datadog and ~20+ other providers. We wanted to share with this community as we know that infra cost management is already a difficult-to-understand, cumbersome task that AI can now help with.
Vantage has a set of logic for organizing costs on top of the base infrastructure level data (which we refer to as Virtual Tags) so that you can organize and allocate spend by service, business unit or function. All of that structured cost data is exposed via MCP to popular LLMs to answer specific questions about your organization.
We put together a quick explainer video on how this works here: https://www.youtube.com/watch?v=69Zt51OSjC0
Docs here: https://docs.vantage.sh/vantage_mcp
I know AI is peak hype right now. But it has definitely changed some of our dev workflows already. So we wanted to find a way to let our customers experiment with how they can use AI to make their cloud cost management work more productive.
The MCP Server acts as a connector between LLMs (right now only Claude, Cursor support it but ChatGPT and Google Gemini coming soon) and your cost and usage data on Vantage which supports 20+ cloud infra providers including AWS, Datadog, Mongo, etc. (You have to have a Vantage account to use it since it's using the Vantage API)
Video demo: https://www.youtube.com/watch?v=n0VP2NlUvRU
Repo: https://github.com/vantage-sh/vantage-mcp-server
It's really impressive how capable the latest-gen models are with an MCP server and an API. So far we have found it useful for:
Ad-Hoc questions: "What's our non-prod cloud spend per engineer if we have 25 engineers"
Action plans: "Find unallocated spend and look for clues how it should be tagged"
Multi-tool workflows: "Find recent cost spikes that look like they could have come from eng changes and look for GitHub PR's merged around the same time" (using it in combination with the GitHub MCP)
Thought I'd share, let me know if you have questions.
We're the team behind Vantage and ec2instances.info and wanted to share a new microsite we're launching. We profiled the number of distinct AWS billing codes across our customer base and found that there are about 60,000 unique ones. We hear all the time that people are confused about the billing codes present in Cost Explorer or the Cost and Usage Report. Think of these as being things like “Requests-Tier1” for S3 or “CW:GMWI-Metrics” for CloudWatch. There is usually really limited resources for determining what these billing codes are and the AWS documentation isn't always fully helpful.
Over the span of the last two months, our team decided to build https://cur.vantage.sh/ - a new microsite for looking up these billing codes and attempting to explain in simplistic terms what these billing codes are and how you're charged. We have written content for a number of services and the site is still in progress but we decided to put it live and start getting feedback. Let us know what you think and if there is anything else that can be helpful on a site like this. Please have mercy as we're trying to share this early to get thoughts.
I'm Co-Founder & CEO of Vantage: a cloud cost observability and optimization platform. We just recently launched support for Github Cost Support within Vantage to view costs for Actions, Shared Storage and Copilot.
We've heard from people that it's been difficult to view GitHub costs; the billing APIs provided insufficient granularity, and the generated CSVs required manual download and review. Often, people did not track their GitHub usage at all throughout the month, and they looked at their bill only at the end of the month. Lastly some people that used self-managed runners, including those that run on Kubernetes or on virtual machines, found it difficult to associate these compute costs with their overall GitHub spend. Allocating build costs by repository, user, or organization was not possible through the GitHub-provided tooling without time-consuming and manual effort.
Users can grant Vantage access to their GitHub billing data by inviting a Vantage-owned email address as a billing administrator to their Enterprise. This billing user allows Vantage to download the detailed usage CSV for maximum billing granularity for GitHub Actions, Storage, and Copilot. For example, customers can see per-user, repository, and workflow-level billing for Actions builds, as well as the detailed usage report for GitHub Enterprise, which enables tracking the license usage of Enterprise members.
Here's a video to see this in action and we'd love to hear your feedback:
https://www.youtube.com/watch?v=MN7RNVo9nc4
(The text below is based upon the linked video above)
I'm Co-Founder & CEO of Vantage: a cloud cost observability platform. Our team put together a quick video showing a new feature we launched that I think this community may be interested in: Network Flow Reports. This gives you the ability to see flows normalized by cost instead of usage for AWS resources.
Knowing that AWS has different rates depending on the source and destination of network traffic, we did the work of ingesting VPC Flow Logs and cross-referencing with the corresponding CUR line-items to be able to show you Network Cost Flows.
I'd love to solicit feedback and see what people think on this!