Hi Infracost team, I’ve encountered an issue where the Infracost GraphQL API does not return Spot pricing for certain i7i instance types (e.g., i7i.xlarge and i7i.2xlarge), even though these Spot prices are available on the AWS EC2 Spot Pricing page (https://aws.amazon.com/ec2/spot/pricing/). This is working fine for

on_demand

instance though Example Query:

query {
  products(
    filter: {
      vendorName: "aws"
      service: "AmazonEC2"
      productFamily: "Compute Instance"
      region: "us-east-1"
      attributeFilters: [
        { key: "instanceType", value: "i7i.2xlarge" }
        { key: "operatingSystem", value: "Linux" }
        { key: "licenseModel", value: "No License required" }
        { key: "tenancy", value: "Shared" }
        { key: "capacitystatus", value: "Used" }
        { key: "preInstalledSw", value: "NA" }
      ]
    }
  ) {
    prices(filter: { purchaseOption: "spot" }) {
      USD
    }
  }
}

what is the reason for this?

#C046GMHQ6NM Hello Team, I'm trying to run infracost from Gitlab pipeline but getting Forbidden method POST url https://pricing.api.infracost.io/graphql , can I just run the tool with API key or do we need the Gitlab integration in order to run it from any CI source Thanks!

Hello, all. I am looking to set up an exceptions file in one of our repos for a FinOps policy. To do this, I believe I need the

policy_id

. I already tried exporting the FinOps policies to CSV, but that does not seem to include the

policy_id

value. Is there any way for me to get the

policy_id

from the InfraCost portal, or will I need to download the repo and run the infracost CLI to get it?

Hi Infracost Team, I am currently facing an issue with the few project cost. Specifically, some environments such as

qa2

and

staging2

are not being detected. Can I configure the autodetect in usgae file eg below:-

version: 0.1
resource_type_default_usage:
  azurerm_app_configuration:
    monthly_additional_requests: 1666666
autodetect:
  env_names:
    - dev
    - qa*
    - staging*
    - prod
    - dr

Hi team!.. I need help pleasee…I am run Infracost with Gitlab App Integrations and comments in PR events, but all projects fails with this error when pipeline run .

No source files found: no files found in directory: /usr/src/app/****/infra/tg_resources/tg_gcp_redis/feature_COPS-7017/****/us-east1/data/prod/data

I use a config.tmpl for generate infracost.yml with projects name and paths, in local the breakdown command work but in Gitlab Pipeline this is currently fail

Hi <!here> i am new to infracost, i tired integrating infracost with GitHub actions but when i try to open a pull request i am not able to see the cost comparison

Hello everyone. I wanted to add infracost to my Atlantis, but I bump into a small problem. We use gitlab and it is in a private network, its domain is not accessible from the outside.

infracost comment gitlab

wants to send me information, but it can't. I have external endpoints in Atlantis, maybe using it it can be bypassed. Thanks in advance!

Hello! Is templating supported with the usage files (e.g.

infracost-usage.yaml

)? Apologies if I missed this somewhere in the documentation.

Hello everyone, I would like to assess the potential maximum cost of my infrastructure on aws per month. What's the best way to simulate that high-usage configuration? I have .tf resources, but would like to assess the potential high-end cost. Is there some kind of a way to generate a usage file from the existing terraform resources with their "maximum" usage capacity? Thanks a lot

Hi guys. How's everybody doing? I just wanted to let you know that I've made a PR on the Infracost docs. We had a link to the Semaphore CI/CD integration (in 3rd party integrations) but the link has changed. The PR updates the URL to the correct one: https://github.com/infracost/docs/pull/748 Sorry if this is not the correct channel for this. I didn't know where to post this.

Hey, I have a pretty general question because I have trouble understanding licensing for Infracost. Can we use infracost API for free as a company? Or will that mean breaching licensing agreements? License in github is Apache so that would normally mean we're good to go but license on infracost site is more strict and I'm confused here.

Dear Infracost Team, I hope you’re doing well. I recently came across Infracost and was truly impressed by the product especially how it brings cost visibility to infrastructure as code, which I find incredibly valuable for DevOps teams. Allow me to introduce myself — my name is Rahul, and I’m a DevOps Engineer and Golang Developer. While exploring open-source projects, I discovered Infracost and was excited to see that it’s open source. I would love to contribute and add value to the project. I’ve joined the Infracost community, but I noticed that some channels are restricted. Could you please guide me on where I can discuss issues, identify suitable beginner-friendly tasks, and understand the best way to start contributing? I’m eager to learn from the community and make meaningful contributions.

Are there any data schemas for the campaign data export? https://www.infracost.io/docs/infracost_cloud/data_export/ I can see the instructions on how to export data, but what is the structure / schema of the exported data? Sample data would also be fine

set the channel topic: ❓For discussions and help with any Infracost issues, also see https://github.com/infracost/infracost/issues

hello team 👋 . first time contributor here. requesting a PR review here: https://github.com/infracost/infracost/pull/3471 relevant issue: https://github.com/infracost/infracost/issues/3472

Questions with regards to Enterprise licenses. I'm thinking about onboarding new users with a self-service request form, approved by manager. If we run out of seats, what can or should we do? 🙂 • call an Infracost API to increase the seats? • contact Infracost support via email? • simply add user in SSO and the total license fee will go up on its own?

Hi Everyone. My project is using terragrunt to provision infrastructure. For pull request automation we also use an Atlantis. In atlantis config we have partion of code that related to infracost:

- run: |
                  # Convert binary plan file to JSON
                  cd $(dirname $PLANFILE)/.terragrunt-cache/*/*
                  terraform show -json $PLANFILE > /tmp/plan.json
                  infracost breakdown --path=/tmp/plan.json \
                    --format=json \
                    --out-file=/tmp/infracost.json
                  INFRACOST_VCS_PROVIDER=gitlab \
                  INFRACOST_VCS_REPOSITORY_URL=<https://gitlab.sportlab.me/betting/infrastructure/terragrunt> \
                  INFRACOST_VCS_TOKEN=$GITLAB_TOKEN \
                  infracost output --path=/tmp/infracost.json \
                    --format=gitlab-comment \
                    --show-skipped

For now we started to put terragrunt cache to single dir, with help of system variable TERRAGRUNT_DOWNLOAD_DIR. As you may see, before we used cd command to move to .terragrunt-cache directory which was located inside terragrunt module. But now we started to locate cache in dedicated directory inside which there are directories named with only hash. Because of that we receive an error:

2025-11-27T14:44:03Z ERROR Could not autodetect any projects from path "/tmp/plan.json"
2025-11-27T14:44:03Z INFO Output saved to /tmp/infracost.json

Maybe someone can help with suggestion how to solve this? Thanks in advance!

Any advice on what to do when a repo gets stuck like this?

Hello, I have a question regarding the Infracost pricing model: • Is the Infracost CI/CD offering available for enterprise customers? • If needed, can we purchase multiple packs to increase the number of calls to the pricing API?

Hello! We have a Terraform mono-repo style terraform project for Azure infrastructure. Each module for a separate Subscription which contains their own modules, some reusable, most not. There are dependencies between each subscription module. We want to get costs for each subscription and in total. We can get it for all resources but if we try to get the cost for a subscription module it will complain about variables not being defined. The problem is that these variables it is complaining about are output from resources in other Subscription modules, these are not values that we can specify manually, at least not easily. Is there a way for us to configure an infracost.yaml config file to look in other modules just for the variable values? Or am I going about this all wrong? The main aim of this is to baseline the cost for each subscription so that we can use this information to configure budgets in Azure.

Heads up —

install.sh

in master appears to be broken against the current

latest

tarball, so CI pipelines using the canonical curl-pipe-sh install are failing. The script ends with:

mv /tmp/infracost-$os-$arch /usr/local/bin/infracost

but the tarball at

<https://infracost.io/downloads/latest/infracost-linux-amd64.tar.gz>

now contains a binary named just

infracost

, with no platform suffix:

$ curl -sL <https://infracost.io/downloads/latest/infracost-linux-amd64.tar.gz> -o /tmp/i.tgz
$ tar tzf /tmp/i.tgz
infracost

Result in CodeBuild logs:

Moving /tmp/infracost-linux-amd64 to /usr/local/bin/infracost ...
mv: cannot stat '/tmp/infracost-linux-amd64': No such file or directory

Checksum validation passes (so the .sha256 was regenerated to match the new archive), and

set -e

doesn't trip on the

tar xzf

, which confirms the archive is structurally fine — just laid out differently than install.sh expects. Looks like the artifacts at

/downloads/latest/

got re-uploaded with a different internal layout without a corresponding install.sh update. Anyone else seeing this? Ally deploys at my company are now broken. This step has always worked for 4.5 years. Happy to open a GitHub issue if it's not already on the radar.

Hey Infracost team It looks like the https://infracost.io/downloads/latest URL is currently broken following the recent changes to the Infracost CLI I also found multiple references to this URL across your repos, for example: https://github.com/search?q=org%3Ainfracost%20https%3A%2F%2Finfracost.io%2Fdownloads%2F&amp;type=code Might be worth updating or redirecting these to the correct location to avoid issues for users relying on the old link. Thanks!

Hey, nice to meet you. I’m a full-stack dev working mainly with TypeScript and Python. A lot of my work is around AI apps lately. Things like LLM integrations, RAG, and agent-style systems. Kind of focused on making them actually useful in real products, not just demos. I usually work with Next.js, Node, and FastAPI. For AI stuff, tools like LangChain or Vercel AI SDK come up a lot. Also using vector DBs like Pinecone or FAISS depending on what fits. For infra, I go with AWS or serverless setups most of the time. Also I handled integrations like Stripe, Supabase, Twilio, stuff like that. I care a lot about scalability and usability. Clean systems that don’t break later on.

Hiya Infracost team, We are looking at using the API only integration and are looking to purchase additional runs, however we noticed that the stripe link in the ui is linked to a recurring charge sku not a usage based sku. We were expecting that the $250 charge would be a 1 time charge for the additional runs, not a recurring $250.

hello! I'm looking into running a POC for Infracost at my company and was wondering what settings are available for GitHub App/Cloud to reduce PR comment noise when there are no significant changes? Features like in issue #3051 available in any kind of offering?

Hey Guys, just looking for some quick assistance with basic Integration with our GH PRs and Private Registry on TFC. I'm getting a 401 error, but the API Key is fine (tested via curl and was able to retrieve all of our modules). I also wrote my own workflow that uses infracost CLI with the same token and it works fine. Am I missing something?

──────────────────────────────────
Project: workspaces
Module path: workspaces/workspace
Errors:
  Failed to load module:
    error retrieving module "app.terraform.io/xxxxxxxxxxxxxxxxx":
      cached error for "app.terraform.io/xxxxxxxxxxxxxxxxx":
        module versions endpoint returned status code 401

PR comment from Infracost App and Test curl screenshots below.

Hi all, the whole EC2

.metal

SKU family is absent from the Cloud Pricing API. Could you add the bare-metal SKUs? Happy to file a GitHub issue if that's the better channel. Thanks!

Hi Infracost team, We use

infracost-diff

(GitHub Action) on PRs to post the cost-diff comment ("Monthly estimate increased by $X..."). We're on the free tier (1,000 runs/month) and have two questions: 1. What exactly happens once we exceed the 1,000 free runs/month? Does the action stop posting the cost-diff comment entirely for the rest of the billing period, or does it degrade gracefully (e.g. only losing the FinOps/tagging-policy and guardrail sections while the basic "Monthly estimate increased/decreased by $X" table still posts)? 2. For our use case, the plain cost-diff comment is what matters most day-to-day — the FinOps policies, tagging checks, guardrails, and Cloud dashboard/leaderboard are nice extras but not essential for us right now. Is there a supported, uncapped way to get just that cost-diff PR comment without it counting against the 1,000 runs/month (or without requiring an Infracost Cloud org/connection at all)? We noticed the CLI was restructured in v2.x —

infracost breakdown

,

infracost diff

, and

infracost comment github

no longer exist, and `infracost scan`/`infracost ci setup` now seem to require an Infracost Cloud org connection. Is there still a standalone, open-source-only path (no Cloud account, no run cap) to compare two Terraform plans and get a cost delta, or has that capability been fully retired ?

good morning, is self-hosting still an option? I was working through the deployment steps under the

Docker compose

section and running into npm build errors.

Good afternoon, I'm trying to setup infracost both within my IDE (VS Code) as well as for a comment within my PRs. Locally I run Windows, VSCode. Our git repos are in Azure DevOps and we use Azure Pipelines. For now I'm only interested in running infracost as the free plan. We are starting up to use Azure, but our costs are not very high yet, but they are expected to rise in the future so we're implementing basic finops measures, which could be extended later on. The terraform projects I'm running are using both modules from the repository as well as from AVM (so online). Not detecting projects In the root of my project I have statically defined my projects like this:

version: 0.1
projects:
  - path: terraform/apps/conn
    name: conn
    terraform_var_files:
      - env/con-production.tfvars
  - path: terraform/apps/kko
    name: kko-startup
    terraform_var_files:
      - env/kko-startup.tfvars
  - path: terraform/apps/mgt
    name: mgt
    terraform_var_files:
      - env/mgt-production.tfvars
  - path: terraform/apps/szg
    name: szg-startup
    terraform_var_files:
      - env/szg-startup.tfvars

It looks like they are recognized during the scan, but not during the PR comment. In the PR comment I see a table with the projects like this: Changed project - Module path DevOpsOrg/DevOpsProject/Azure - terraform/apps/kko DevOpsOrg/DevOpsProject/Azure - terraform/apps/szg DevOpsOrg/DevOpsProject/Azure-production - terraform/apps/mgt DevOpsOrg/DevOpsProject/Azure/terraform/apps/kko - terraform/apps/kko DevOpsOrg/DevOpsProject/Azure/terraform/apps/mgt-production - terraform/apps/mgt DevOpsOrg/DevOpsProject/Azure/terraform/apps/szg - terraform/apps/szg The costs are being moved around, maybe that is because this is the first run, but it looks like the projects are being autodetected, which I also see in the output of the tasks: 2026-08-18T092928Z INFO Autodetected 4 Terraform projects across 4 root modules - Local scan in VS Code does not work When I run infracost scan locally all projects either error out with the error below, or the scan runs for hours and never finishes: 2026-08-18 110910.674 [error] time=2026-08-18T110910.671+02:00 level=ERROR msg="analyzeFullScan: project scan failed" name=szg-startup error="parsing: rpc error: code = DeadlineExceeded desc = context deadline exceeded" elapsed=1m0.3333715s In the meantime, the extension says it's scanning the workspace, which does not end. It just keeps on scanning without any output or error message. I've ran Infracost doctor: Infracost Doctor 2.16.2 - running 7 checks Authentication ✔️ Credentials found ✔️ Token valid ✔️ Organization accessible "hagaziekenhuis" ✔️ API reachable (385 ms) CLI ✔️ Version 2.16.2 (latest) Configuration ✔️ Config file valid ✔️ Default org set (hagaziekenhuis) ✔️ 7 passed An I've also cleaned up the cache: infracost cache clear ✓ Cache cleared. After clearing the cache, when running command below I get popups for GitHub, even though this is Azure Repos: infracost breakdown --path=. --format=json --out-file=./breakdown.json !

infracost breakdown

is deprecated, running

infracost scan

instead. ✔️ Plugins up to date ⡿ Scanning... ⣾ Scanning... ⣾ Scanning... ⣾ Scanning... ⢿ Scanning... - I also get messages in vscode that the ls has a new version, but when trying an update it says: 2026-08-18 115659.269 [error] time=2026-08-18T115659.268+02:00 level=ERROR msg="method error" method=infracost/update duration=1.0762821s error="replacing binary: rename C:\\Users\\sjoer\\.vscode\\extensions\\infracost.infracost-0.4.17-win32-x64\\bin\\.infracost-ls-update-1706734103 C\\Users\\sjoer\\.vscode\\extensions\\infracost.infracost 0.4.17 win32 x64\\bin\\infracost ls.exe Access is denied." - And a final additional question, is the cli within vscode included in the free edition? Because I don't see anything about that on the pricing page.