Extended FAQ

This expanded FAQ covers everything users ask most often.

---

❓ What is Prompt Tornado?

Prompt Tornado is a multi-model AI orchestration platform that automates complex workflows (text, research, images, code, etc.) from a single prompt.

---

❓ How does it choose which model to use?

Using intelligent routing + a 3-tier fallback system:

• Best model for the task

• Secondary fallback

• Tertiary / Lightweight fallback

I have striven to diligently combine LLM-as-a-Judge and my own judgment / rankings to evaluate which models tend to work best for particular task types. If for any reason you are not satisfied with the results, you can always try again (i.e. run another workflow) by specifying which supported models you'd prefer were used for each step in your workflow.

One more note. We have intentionally chosen which AI providers and models our platform supports (and which we do not). Some providers and models simply do not meet our requirements (sometimes because of immense technical integration challenges, and sometimes due to a lack of alignment on human values preferences). We reserve the right to choose which providers we support and which we do not (but you may submit feedback or recommendations for providers you want us to support in the future at any time).

We try to avoid working with providers or models that we believe are not aligned with the interests of our users. We may deem a certain model as too unsafe, promoting toxicity/bias or objectively inaccurate information, or even posing a national security risk / possibly collecting or disclosing our users' information without their express consent or with the intent to undermine American interests.

---

❓ Can I specify which model I want used?

Yes. Just say:

“Use Claude Haiku for this.”

If the model is unsupported, Prompt Tornado uses normal routing.

---

❓ What data do you store?

We store:

• Prompt text

• Model usage

• Outputs

• Workflow metadata

You can request deletion of your account at any time.

---

❓ How does research work?

Research tasks automatically use Perplexity Sonar Reasoning Pro for:

• Live & Up-to-Date internet results

• Reliable sources with accurate info

• Synthesis of research insights

---

❓ Why did my workflow run multiple steps when I only wanted one?

Your prompt likely implied (or could reasonably be misinterpreted as) requesting multiple tasks (or possibly was written in a way that could have been clearer).

Example:

“Write an article and then generate audio of it.”

---

❓ Does Prompt Tornado support uploading files (text, image, and/or video) to analyze, edit, or provide context?

Not yet. This is something we are considering supporting in the future.

---

❓ What’s coming next?

• Support for more task types (i.e. video generation, data analysis)

• Custom workflow templates

• API access for developers

---