# Using two different models for the same application

**URL:** <https://forum.rasa.com/t/using-two-different-models-for-the-same-application/2467>\
**Category:** Rasa Open Source\
**Created:** [November 6, 2018, 10:07am UTC](https://forum.rasa.com/t/using-two-different-models-for-the-same-application/2467 "2018-11-06T10:07:56Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![LeonPaul237](https://avatars.discourse-cdn.com/v4/letter/l/439d5e/32.png) [@LeonPaul237](https://forum.rasa.com/u/LeonPaul237)\
**Post date:** [November 6, 2018, 10:07am UTC](https://forum.rasa.com/t/using-two-different-models-for-the-same-application/2467/1 "2018-11-06T10:07:56Z")

</div>

Hi,

I am currently working on a framework that involves certain business use case intents as well as a set of small talk intents. The issue, is that some of the small talk dialogues conflict with my business use cases when integrated.

Is there a way for me to have two seperately trained models, one for the business use cases and the other for small talk exclusively?

Apologies if this query has already been answered, I’m new to developing with RASA and any advice would be greatly appreciated!!

Thanks in advance!!

---

<div class="post-metadata">

**Author:** ![ErikTromp](https://dub1.discourse-cdn.com/flex013/user_avatar/forum.rasa.com/eriktromp/32/993_2.png) [@ErikTromp](https://forum.rasa.com/u/ErikTromp)\
**Post date:** [November 6, 2018, 10:37am UTC](https://forum.rasa.com/t/using-two-different-models-for-the-same-application/2467/2 "2018-11-06T10:37:10Z")

</div>

So what stops you from training the two models and just running two instances of Rasa on two different dockers or ports or whatever?

---

<div class="post-metadata">

**Author:** ![LeonPaul237](https://avatars.discourse-cdn.com/v4/letter/l/439d5e/32.png) [@LeonPaul237](https://forum.rasa.com/u/LeonPaul237)\
**Post date:** [November 6, 2018, 10:50am UTC](https://forum.rasa.com/t/using-two-different-models-for-the-same-application/2467/3 "2018-11-06T10:50:07Z")

</div>

I could try that approach, but how do I control which Interpreter gets used based on the intent? Or is that taken care off by the Stack itself?

---

<div class="post-metadata">

**Author:** ![ErikTromp](https://dub1.discourse-cdn.com/flex013/user_avatar/forum.rasa.com/eriktromp/32/993_2.png) [@ErikTromp](https://forum.rasa.com/u/ErikTromp)\
**Post date:** [November 15, 2018, 11:08am UTC](https://forum.rasa.com/t/using-two-different-models-for-the-same-application/2467/4 "2018-11-15T11:08:09Z")

</div>

Ah you want to run two rasa models within the same chat… I am wondering what your usage scenario is then - what prevents you from modeling the dialogue as one instance?