# Struggling to choose the right AI model for your project?

**URL:** <https://community.postman.com/t/struggling-to-choose-the-right-ai-model-for-your-project/76422>\
**Category:** API Pulse\
**Tags:** eval, ai, agentic-ai\
**Created:** [March 4, 2025, 7:38pm UTC](https://community.postman.com/t/struggling-to-choose-the-right-ai-model-for-your-project/76422 "2025-03-04T19:38:42Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![sterlingchin](https://sea2.discourse-cdn.com/getpostman/user_avatar/community.postman.com/sterlingchin/32/31421_2.png) [@sterlingchin](https://community.postman.com/u/sterlingchin)\
**Post date:** [March 4, 2025, 7:38pm UTC](https://community.postman.com/t/struggling-to-choose-the-right-ai-model-for-your-project/76422/1 "2025-03-04T19:38:42Z")

</div>

Ever found yourself wondering which LLM would work best for your specific needs? In our latest video, I walk you through TWO powerful methods to evaluate AI models like GPT, Claude, Gemini, and others using Postman!

The first method uses Collection Runner to benchmark multiple models simultaneously, comparing performance metrics like token usage, response time, and content length all at once. Perfect for data-driven developers who need comprehensive testing.

 ![Postman Collection Runner comparing multiple AI models, displaying token usage, response time, and content length for benchmarking.](https://us1.discourse-cdn.com/getpostman/original/3X/c/2/c20136ec6335c9bc137bbf8d077e85d5f1806d43.jpeg)

What’s your go-to method for comparing AI models? Do you prefer quantitative metrics or qualitative response evaluation?

The second approach uses Postman Flows for real-time, interactive comparisons. This visual method lets you see exactly how different models respond to the same prompt simultaneously, with an AI evaluator determining which response is best based on your custom criteria.

 ![Postman Flows showing real-time AI model comparisons, visually displaying responses to the same prompt with an AI evaluator selecting the best one.](https://us1.discourse-cdn.com/getpostman/original/3X/2/e/2ea624584e540033dff01445f11097942659c6a8.jpeg)

Both the Collection Runner and Flow examples from this tutorial are available in our public workspace! Have you tried either of these approaches before? Which metrics matter most to you when selecting an AI model?

📺 Watch the full tutorial here to see both methods in action:

[![](https://img.youtube.com/vi/S_u_LR7Gce4/maxresdefault.jpg "Evaluate LLM Performance in Postman") ](https://www.youtube.com/watch?v=S_u_LR7Gce4)

---

<div class="post-metadata">

**Author:** ![allenheltondev](https://sea2.discourse-cdn.com/getpostman/user_avatar/community.postman.com/allenheltondev/32/31738_2.png) [@allenheltondev](https://community.postman.com/u/allenheltondev)\
**Post date:** [March 4, 2025, 11:04pm UTC](https://community.postman.com/t/struggling-to-choose-the-right-ai-model-for-your-project/76422/2 "2025-03-04T23:04:57Z")

</div>

Love everything about this!

My favorite thing of all just might be that 90’s sitcom style thumbnail for the video 🤣