# Text to Image Reliability Concern

**URL:** <https://community.pickaxe.co/t/text-to-image-reliability-concern/2472>\
**Category:** General\
**Created:** [November 7, 2024, 4:42pm UTC](https://community.pickaxe.co/t/text-to-image-reliability-concern/2472 "2024-11-07T16:42:07Z")\
**Posts on this page:** 6\
**Page:** 1

<div class="post-metadata">

**Author:** ![chadvmm](https://avatars.discourse-cdn.com/v4/letter/c/ecd19e/32.png) [@chadvmm](https://community.pickaxe.co/u/chadvmm)\
**Post date:** [November 7, 2024, 4:42pm UTC](https://community.pickaxe.co/t/text-to-image-reliability-concern/2472/1 "2024-11-07T16:42:07Z")

</div>

![Screenshot 2024-11-07 at 11.41.09 AM](https://canada1.discourse-cdn.com/flex004/uploads/pickaxeproject/original/2X/0/0a9c7e3a9bcb4793e02343b66b3e267a71848b9d.png)

My text to image seems VERY hit or miss. Very unreliable. Not sure what else to do  
@admin_mike

---

<div class="post-metadata">

**Author:** ![nathanielmhld](https://yyz2.discourse-cdn.com/flex004/user_avatar/community.pickaxe.co/nathanielmhld/32/11_2.png) [@nathanielmhld](https://community.pickaxe.co/u/nathanielmhld)\
**Post date:** [November 8, 2024, 12:40am UTC](https://community.pickaxe.co/t/text-to-image-reliability-concern/2472/2 "2024-11-08T00:40:04Z")

</div>

It looks like you’re trying to generate images that could be flagging the model’s content filters. Depending on what model you’re using, they could be very strict. What model are you using?

---

<div class="post-metadata">

**Author:** ![chadvmm](https://avatars.discourse-cdn.com/v4/letter/c/ecd19e/32.png) [@chadvmm](https://community.pickaxe.co/u/chadvmm)\
**Post date:** [November 8, 2024, 12:51am UTC](https://community.pickaxe.co/t/text-to-image-reliability-concern/2472/3 "2024-11-08T00:51:49Z")

</div>

Interesting. Dall-E, chat gpt

---

<div class="post-metadata">

**Author:** ![nathanielmhld](https://yyz2.discourse-cdn.com/flex004/user_avatar/community.pickaxe.co/nathanielmhld/32/11_2.png) [@nathanielmhld](https://community.pickaxe.co/u/nathanielmhld)\
**Post date:** [November 8, 2024, 12:57am UTC](https://community.pickaxe.co/t/text-to-image-reliability-concern/2472/4 "2024-11-08T00:57:36Z")

</div>

We moved our default image generation from Dall-E to Flux specifically because Dall-E constantly flags requests all the time. If you’re asking for something about “women’s pelvic health”, even though it’s totally innocuous, I could easily see Dall-E flagging it. Use Flux or a different model.

---

<div class="post-metadata">

**Author:** ![chadvmm](https://avatars.discourse-cdn.com/v4/letter/c/ecd19e/32.png) [@chadvmm](https://community.pickaxe.co/u/chadvmm)\
**Post date:** [November 8, 2024, 1:24am UTC](https://community.pickaxe.co/t/text-to-image-reliability-concern/2472/5 "2024-11-08T01:24:15Z")

</div>

Interesting! Is that better than midjourney in your opinion?

I’m so tired of the 3 arms, 3 legs, abstract positions these images come Back with

---

<div class="post-metadata">

**Author:** ![nathanielmhld](https://yyz2.discourse-cdn.com/flex004/user_avatar/community.pickaxe.co/nathanielmhld/32/11_2.png) [@nathanielmhld](https://community.pickaxe.co/u/nathanielmhld)\
**Post date:** [November 8, 2024, 1:37am UTC](https://community.pickaxe.co/t/text-to-image-reliability-concern/2472/6 "2024-11-08T01:37:35Z")

</div>

Flux comes free with Pickaxe and is easier to set up than Midjourney. But many people like Midjourney better, the setup process is pretty involved though.