# Needing to purchase API Tokens

**URL:** <https://forum.inductiveautomation.com/t/needing-to-purchase-api-tokens/117772>\
**Category:** Ignition 2027 Early Access\
**Created:** [November 6, 2026, 2:32pm UTC](https://forum.inductiveautomation.com/t/needing-to-purchase-api-tokens/117772 "2026-11-06T14:32:54Z")\
**Posts on this page:** 6\
**Page:** 1

<div class="post-metadata">

**Author:** ![Curlyandshemp](https://avatars.discourse-cdn.com/v4/letter/c/bcef8e/32.png) [@Curlyandshemp](https://forum.inductiveautomation.com/u/Curlyandshemp)\
**Post date:** [November 6, 2026, 2:32pm UTC](https://forum.inductiveautomation.com/t/needing-to-purchase-api-tokens/117772/1 "2026-11-06T14:32:54Z")

</div>

I am trying Ignition - 2027 Beta and have OpenAI set up and configured properly. When I try to use ASK AI within Ignition and run a test AI query in the Script Console, I get an error message "..no credits remaining. Add credits to continue using..."

I am now retired and am just trying to keep update with the Inductive Automation advancements. Been using Ignition since the Blue Ravine days with the funky Dvorak keyboards.

As a "hobbyist" is there a way , or a decent AI model that can be used without having to lay out cash ? I totaly get the AI companies need to make money.

Thanks

Ian Phillips

---

<div class="post-metadata">

**Author:** ![Noah\_Burnette](https://sea1.discourse-cdn.com/inductiveautomation/user_avatar/forum.inductiveautomation.com/noah_burnette/32/83777_2.png) [@Noah\_Burnette](https://forum.inductiveautomation.com/u/Noah_Burnette)\
**Post date:** [November 6, 2026, 2:59pm UTC](https://forum.inductiveautomation.com/t/needing-to-purchase-api-tokens/117772/2 "2026-11-06T14:59:38Z")

</div>

Hey Ian,

I haven't tested but you can configure the OpenRouter model provider and use a free model (obviously with rate limits and will probably use your session to train 😃).

> **[Catalyst Core | Ignition User Manual](https://docs.inductiveautomation.com/docs/2027/ignition-modules/catalyst/catalyst-core)**
>
> The Catalyst Core module is bundled with the Ignition platform, and is required to use the Catalyst Agent module.

> **[Compare AI Models: Pricing, Context & Benchmarks | OpenRouter](https://openrouter.ai/models?max_price=0)**
>
> Compare 600+ LLMs from OpenAI, Anthropic, Google, Meta and more — pricing, context length, and benchmarks side by side, all through one API.

---

<div class="post-metadata">

**Author:** ![KathyApplebaum](https://sea1.discourse-cdn.com/inductiveautomation/user_avatar/forum.inductiveautomation.com/kathyapplebaum/32/7261_2.png) [@KathyApplebaum](https://forum.inductiveautomation.com/u/KathyApplebaum)\
**Post date:** [November 6, 2026, 3:16pm UTC](https://forum.inductiveautomation.com/t/needing-to-purchase-api-tokens/117772/3 "2026-11-06T15:16:23Z")

</div>

I've used [Ollama](https://ollama.com/) for testing our AI module and it gave decent results. Set up was pretty straight forward. NAYY

---

<div class="post-metadata">

**Author:** ![michael.flagler](https://sea1.discourse-cdn.com/inductiveautomation/user_avatar/forum.inductiveautomation.com/michael.flagler/32/26272_2.png) [@michael.flagler](https://forum.inductiveautomation.com/u/michael.flagler)\
**Post date:** [November 7, 2026, 4:48am UTC](https://forum.inductiveautomation.com/t/needing-to-purchase-api-tokens/117772/4 "2026-11-07T04:48:47Z")

</div>

What model did you use with Ollama? Curious which one works the best for Ignition.

---

<div class="post-metadata">

**Author:** ![KathyApplebaum](https://sea1.discourse-cdn.com/inductiveautomation/user_avatar/forum.inductiveautomation.com/kathyapplebaum/32/7261_2.png) [@KathyApplebaum](https://forum.inductiveautomation.com/u/KathyApplebaum)\
**Post date:** [November 9, 2026, 3:54pm UTC](https://forum.inductiveautomation.com/t/needing-to-purchase-api-tokens/117772/5 "2026-11-09T15:54:47Z")

</div>

It's been a couple of months so I don't recall, but it was probably just whatever the default was.

---

<div class="post-metadata">

**Author:** ![osolorzano](https://sea1.discourse-cdn.com/inductiveautomation/user_avatar/forum.inductiveautomation.com/osolorzano/32/6432_2.png) [@osolorzano](https://forum.inductiveautomation.com/u/osolorzano)\
**Post date:** [November 11, 2026, 6:00pm UTC](https://forum.inductiveautomation.com/t/needing-to-purchase-api-tokens/117772/6 "2026-11-11T18:00:24Z")

</div>

In QA, the local model we've used the most is qwen3:4b as it was small, can do tool calls, and generated decent results. It can be really slow however and can require some finagling with the Max Context Tokens if you don't have much memory available. It typically required going back and forth with some number of tokens available, running a session call from somewhere, and running ollama ps to make sure it's still [running 100% GPU](https://docs.ollama.com/faq#how-can-i-tell-if-my-model-was-loaded-onto-the-gpu) otherwise it spins its wheels.
