# Announcing AI credentials

**URL:** https://meta.discourse.org/t/announcing-ai-credentials/397408
**Category:** Announcements
**Tags:** ai
**Created:** [March 2, 2026, 4:35am UTC](https://meta.discourse.org/t/announcing-ai-credentials/397408 "2026-03-02T04:35:21Z")
**Posts on this page:** 4
**Page:** 1

<div class="post-metadata">

### Author: ![sam](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/sam/32/102149_2.png) [@sam](https://meta.discourse.org/u/sam)
#### Post date: [March 2, 2026, 4:35am UTC](https://meta.discourse.org/t/announcing-ai-credentials/397408/1 "2026-03-02T04:35:21Z")

</div>

We just completed a large refactor that shifted all credentials Discourse AI uses (tools/embeddings and LLMs) to a central store:

 ![image](https://global.discourse-cdn.com/meta/original/4X/7/2/1/7216f58df4df7db990594926f9a50fcf49048d22.png)

This solves a long-standing pain point around tool distribution (you can now easily distribute tools that require secrets). It also makes it significantly easier to rotate credentials and add new models.

Hope you all enjoy it.

Learn more at:

> [@Managing AI credentials](https://meta.discourse.org/t/managing-ai-credentials/397407):
>
> bookmark This guide covers creating, managing, and using AI credentials (shared secrets) across LLMs, embedding models, and custom AI tools in the Discourse AI plugin. person_raising_hand Required user level: Administrator Summary AI credentials provide a centralized, secure way to manage authentication secrets — such as API keys — across your Discourse AI configuration. Instead of pasting raw API keys into each LLM model, embedding definition, or custom tool individually, you create a c…

---

<div class="post-metadata">

### Author: ![Lilly](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/lilly/32/575047_2.png) [@Lilly](https://meta.discourse.org/u/Lilly)
#### Post date: [March 2, 2026, 10:53am UTC](https://meta.discourse.org/t/announcing-ai-credentials/397408/2 "2026-03-02T10:53:03Z")

</div>

i saw this over the weekend after updating one of my instances - so much better now! i was updating my gemini models and rotating api keys and this made it way easier to manage. thank you Sam!

---

<div class="post-metadata">

### Author: ![Lilly](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/lilly/32/575047_2.png) [@Lilly](https://meta.discourse.org/u/Lilly)
#### Post date: [March 2, 2026, 10:58am UTC](https://meta.discourse.org/t/announcing-ai-credentials/397408/3 "2026-03-02T10:58:05Z")

</div>

i just noticed that the mobile viewport size needs some loving 🙂

 ![IMG_5010](https://global.discourse-cdn.com/meta/original/4X/7/d/7/7d7a73777f1fab337f559edf73bed4dfcaf60686.jpeg)

---

<div class="post-metadata">

### Author: ![derek](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/derek/32/458751_2.png) [@derek](https://meta.discourse.org/u/derek)
#### Post date: [March 2, 2026, 4:30pm UTC](https://meta.discourse.org/t/announcing-ai-credentials/397408/5 "2026-03-02T16:30:50Z")

</div>

Thanks for the heads up @Lilly. It should stack the LLM labels on mobile now:

 ![Screenshot 2026-03-02 at 7.55.56 AM](https://global.discourse-cdn.com/meta/original/4X/5/e/4/5e4a58a4c680e0fd8fda9a461bbed8015bda2524.png)