# How does chunking work in OpenAI Assistant API's vector store?

**URL:** <https://community.openai.com/t/how-does-chunking-work-in-openai-assistant-apis-vector-store/1141956>\
**Category:** API\
**Tags:** chatgpt, assistants-api\
**Created:** [March 13, 2025, 4:05am UTC](https://community.openai.com/t/how-does-chunking-work-in-openai-assistant-apis-vector-store/1141956 "2025-03-13T04:05:32Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![harijayaraman05](https://sea2.discourse-cdn.com/openai1/user_avatar/community.openai.com/harijayaraman05/32/419665_2.png) [@harijayaraman05](https://community.openai.com/u/harijayaraman05)\
**Post date:** [March 13, 2025, 4:05am UTC](https://community.openai.com/t/how-does-chunking-work-in-openai-assistant-apis-vector-store/1141956/1 "2025-03-13T04:05:32Z")

</div>

Specifically, I have a doubt—chunking is typically set to a fixed size for all files, without considering the actual content size. However, does OpenAI dynamically determine the chunk size based on the file size, or is it always fixed? Could you provide some insights into this?

---

<div class="post-metadata">

**Author:** ![shounakswipe](https://sea2.discourse-cdn.com/openai1/user_avatar/community.openai.com/shounakswipe/32/511442_2.png) [@shounakswipe](https://community.openai.com/u/shounakswipe)\
**Post date:** [March 13, 2025, 5:22am UTC](https://community.openai.com/t/how-does-chunking-work-in-openai-assistant-apis-vector-store/1141956/2 "2025-03-13T05:22:44Z")

</div>

You can modify the chunking according to your own strategy while you are ingesting the document.