# GLB model too heavy

**URL:** <https://discourse.threejs.org/t/glb-model-too-heavy/38988>\
**Category:** Questions\
**Tags:** models, gltf\
**Created:** [June 7, 2022, 6:42pm UTC](https://discourse.threejs.org/t/glb-model-too-heavy/38988 "2022-06-07T18:42:39Z")\
**Posts on this page:** 5\
**Page:** 1

<div class="post-metadata">

**Author:** ![IssamMerikhi](https://yyz1.discourse-cdn.com/flex035/user_avatar/discourse.threejs.org/issammerikhi/32/25238_2.png) [@IssamMerikhi](https://discourse.threejs.org/u/IssamMerikhi)\
**Post date:** [June 7, 2022, 6:42pm UTC](https://discourse.threejs.org/t/glb-model-too-heavy/38988/1 "2022-06-07T18:42:39Z")

</div>

Hello guys. My question is pretty simple, I load 4 models and some texture for my website but the first loading is too long. You can take a look : [https://issammerikhi.vercel.app/](https://issammerikhi.vercel.app/)

Do you know a simple and efficient way to reduce my model size ?

Thanks for helping

---

<div class="post-metadata">

**Author:** ![drcmda](https://avatars.discourse-cdn.com/v4/letter/d/91b2a8/32.png) [@drcmda](https://discourse.threejs.org/u/drcmda)\
**Post date:** [June 7, 2022, 8:37pm UTC](https://discourse.threejs.org/t/glb-model-too-heavy/38988/2 "2022-06-07T20:37:10Z")

</div>

- always reduce polycount in blender, has lots of cleaning options, removal of verts by distance etc

now, go through processing, for instance:

- [http://gltf.report](http://gltf.report)
- npx gltf-pipeline -i model.gtlf -o model.glb --draco.compressionLevel=10
- npx gltfpack -i model.gltf -o model.glb -cc -si 1

don’t forget to compress and resize the textures, which are often the culprit. i have brought down 300mb gltfs to 1-2-3mb with these tools.

---

<div class="post-metadata">

**Author:** ![IssamMerikhi](https://yyz1.discourse-cdn.com/flex035/user_avatar/discourse.threejs.org/issammerikhi/32/25238_2.png) [@IssamMerikhi](https://discourse.threejs.org/u/IssamMerikhi)\
**Post date:** [June 9, 2022, 7:34am UTC](https://discourse.threejs.org/t/glb-model-too-heavy/38988/3 "2022-06-09T07:34:35Z")

</div>

Ok thanks for the response. I’m trying right now.

At first the file looks smaller.

 ![image](https://canada1.discourse-cdn.com/flex035/uploads/threejs/original/3X/a/f/affcba71935aa4755307ba38c04b98867d8fbcce.png)

But I guess I need Draco compressor to continue

 ![image](https://canada1.discourse-cdn.com/flex035/uploads/threejs/original/3X/6/2/62832a9f5e51395aec8cd28cdad01d361bef86f0.png)

I’ll let you know 🙂

---

<div class="post-metadata">

**Author:** ![JuanP](https://yyz1.discourse-cdn.com/flex035/user_avatar/discourse.threejs.org/juanp/32/25869_2.png) [@JuanP](https://discourse.threejs.org/u/JuanP)\
**Post date:** [June 9, 2022, 8:55am UTC](https://discourse.threejs.org/t/glb-model-too-heavy/38988/4 "2022-06-09T08:55:06Z")

</div>

Hi @IssamMerikhi

The half of the size of your glb are texture/ png images. Try to use webp as versions of the texture and probably you will save 10/15mb. Reduce the size of the texture could help too.

I don’t know any tool that can do that automatically, you should process images one by one and change the reference manually in the gltf. Programatically you can do it with sharp and gltf-transform, with GUIs you can use photoshop, Gimp to save the file as webp and edit the gltf manually in a text editor.

---

<div class="post-metadata">

**Author:** ![jaggss](https://yyz1.discourse-cdn.com/flex035/user_avatar/discourse.threejs.org/jaggss/32/45834_2.png) [@jaggss](https://discourse.threejs.org/u/jaggss)\
**Post date:** [March 17, 2024, 2:51pm UTC](https://discourse.threejs.org/t/glb-model-too-heavy/38988/5 "2024-03-17T14:51:36Z")

</div>

I was going through your website, looks pretty cool. Have you tried [https://optimizeglb.com/](https://optimizeglb.com/). This was working well for me and its very easy.