---
title: "Discussion on \"Prefill vs. decode: Why the same model feels fast in one product and slow in another\" | Hashnode"
description: "Discussion on \"Prefill vs. decode: Why the same model feels fast in one product and slow in another\". There's a question that comes up a lot when teams start measuring AI performance.\n\"Why does our model feel slow? We're using the same model as everyone else.\"\nThe answer almost always lives in the sam"
url: "https://hashnode.com/posts/prefill-vs-decode-why-the-same-model-feels-fast-in-one-product-and-slow-in-another/6a5024d761270a6315fb4aaf"
---

[Feed](https://hashnode.com/feed) /

Discussion

# [Prefill vs. decode: Why the same model feels fast in one product and slow in another](https://leveragepoints.io/prefill-vs-decode-why-the-same-model-feels-fast-in-one-product-and-slow-in-another?utm_source=hashnode&utm_medium=feed)

There's a question that comes up a lot when teams start measuring AI performance. "Why does our model feel slow? We're using the same model as everyone else." The answer almost always lives in the sam

Jul 9 · 4 min read · leveragepoints.io  
Read the article

## Written by

[Abdulwaheed A A](https://hashnode.com/@habiz)

@habiz

  
[View profile](https://hashnode.com/@habiz)

Building Products

I lead the strategy and delivery of products and systems that solve real problems, support commercial goals and scale across teams and markets. My work spans platform foundations, user-facing experiences, commercial tools and developer products, and I adapt my approach to the needs of each domain. My experience covers early-stage, scale-up and enterprise product environments, combining hands-on delivery with team leadership and award-winning work recognised across several innovative organisations.

## Responses 2

Markdown supported · ⌘↵ to post

[Julian Neagu](https://hashnode.com/@julianneagulab)

[Jul 11](https://hashnode.com/posts/prefill-vs-decode-why-the-same-model-feels-fast-in-one-product-and-slow-in-another/6a5024d761270a6315fb4aaf/comment/6a52259c9e6668632488103a)

A lot of people blame the model when the real issue is how the feature uses it. Seen this a few times with AI tools. The same model can feel totally different depending on the flow.

[Abdulwaheed A A](https://hashnode.com/@habiz)

[Jul 11](https://hashnode.com/posts/prefill-vs-decode-why-the-same-model-feels-fast-in-one-product-and-slow-in-another/6a5024d761270a6315fb4aaf/comment/6a52259c9e6668632488103a#reply-6a526344dd362619795372cd)

If most people understand what's happening behind the hood and where where there flow sits, they should be able to optimise it better for a good experience and performance improvement