If you’re building applications with Hugging Face models or Inference Endpoints, you’ve probably tested your APIs with GUI tools like Postman.
That works well during development, but as projects grow, we found it more practical to move API testing into a CLI-based workflow that integrates with automation.
While comparing different options, this article was a useful starting point:
Here’s the workflow we’ve been experimenting with.
Instead of relying only on manual requests in a GUI, organize your API definitions so they’re easy to reuse across development and deployment.
CLI-based tools make it much easier to automate API validation and integrate tests into scripts and CI/CD pipelines.
We’re currently evaluating Apidog CLI, since it lets us execute API test scenarios directly from the terminal and fits naturally into automated workflows.
apidog login
apidog run
Before deploying changes to an AI service or inference endpoint, run the same API tests locally and in your CI pipeline.
This helps catch authentication, schema, or response issues before they reach production.
A CLI-first approach makes it easier for developers, QA engineers, and automation pipelines to run the same API validation without depending on a desktop application.
We’re exploring CLI alternatives to traditional API clients like Postman, especially for AI projects where APIs evolve quickly and automated validation is becoming increasingly important.
I’d be interested to hear what other Hugging Face developers are using.