# PYNQ-DPU without use .xmodel

**URL:** <https://discuss.pynq.io/t/pynq-dpu-without-use-xmodel/4040>\
**Category:** Support\
**Created:** [April 14, 2022, 5:37pm UTC](https://discuss.pynq.io/t/pynq-dpu-without-use-xmodel/4040 "2022-04-14T17:37:49Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![Nagendra](https://sea1.discourse-cdn.com/flex019/user_avatar/discuss.pynq.io/nagendra/32/2653_2.png) [@Nagendra](https://discuss.pynq.io/u/Nagendra)\
**Post date:** [April 14, 2022, 5:37pm UTC](https://discuss.pynq.io/t/pynq-dpu-without-use-xmodel/4040/1 "2022-04-14T17:37:49Z")

</div>

Hi,

I want to run some deep-learning algorithms on the DPU using a.bit file instead. xmodel Is it possible to do that?

Thank you

---

<div class="post-metadata">

**Author:** ![skalade](https://sea1.discourse-cdn.com/flex019/user_avatar/discuss.pynq.io/skalade/32/686_2.png) [@skalade](https://discuss.pynq.io/u/skalade)\
**Post date:** [April 14, 2022, 10:44pm UTC](https://discuss.pynq.io/t/pynq-dpu-without-use-xmodel/4040/2 "2022-04-14T22:44:14Z")

</div>

Hi there,

If you want to use the DPU you will need the Vitis AI runtime (VART) API. In order to use VART, you need a .xmodel for the DPU to be able to run your model. I’m not aware of communicating with the DPU without the VART API, but maybe you can ask the Vitis AI maintainers.

Thanks  
Shawn

---

<div class="post-metadata">

**Author:** ![Nagendra](https://sea1.discourse-cdn.com/flex019/user_avatar/discuss.pynq.io/nagendra/32/2653_2.png) [@Nagendra](https://discuss.pynq.io/u/Nagendra)\
**Post date:** [April 15, 2022, 6:11am UTC](https://discuss.pynq.io/t/pynq-dpu-without-use-xmodel/4040/3 "2022-04-15T06:11:55Z")

</div>

Hi @skalade

Thank you for your suggestions.