*Thisarticle is based on a YouTube video by Sophie Hundertmark, an expert in the use of artificial intelligence with a focus on chatbots and strategic AI applications in companies and public institutions. Sophie is a researcher and lecturer at the Lucerne University of Applied Sciences and Arts and is doing her doctorate in Conversational AI at the University of Fribourg. The blog text was created using a custom GPT model that was trained on Sophie’s video content, language style and expertise. The result is well-founded, up-to-date articles based on Sophie Hundertmark‘s own expertise.
You can find the link to the video at the end of this article.
The range of AI language models is growing rapidly. With every new version of ChatGPT, Claude, or Gemini, the same question arises: Which model is actually the right one?
The good news is: In most cases, you don’t really have to worry about it that much.
Instead of spending a long time thinking about the perfect model, you should just get started. For many everyday tasks, the standard models are perfectly adequate. Only when your needs become more complex is it worth switching to more powerful models.
In this post, I’ll show you when simple models are sufficient and in which situations more powerful language models are really worth it.
The most important rule: Don’t think about it for too long
Many users spend more time choosing a language model than on the actual task.
The following applies:
- It’s easy to get started.
- Test different models.
- Gain your own experience.
The differences between the models are often smaller than expected—especially for simple tasks.
Simple tasks don’t require a high-performance model
The standard models are perfectly adequate for everyday inquiries.
These include, for example:
- Small Talk in a Foreign Language
- simple questions
- short texts
- Summaries
- Translations
For example, if you want to practice Italian or have a short text translated, even a basic model can deliver very good results.
The big advantage:
- fast response times
- low resource consumption
- more than enough for standard tasks
When using ChatGPT, I often use GPT-5.5 for tasks like this because it works quickly and is more than sufficient for everyday queries.
Translation is one of the simplest tasks for AI
Translations, too, generally do not require a complex conceptual framework.
A simple prompt like:
“Translate this text into English.”
Even with the fast standard models, it delivers excellent results.
In most cases, it’s not worth thinking too deeply about this. The model does not need to analyze complex relationships; it simply needs to translate the text accurately.
Moderately difficult problems benefit from a little more time to think
As soon as multiple criteria need to be taken into account, a more powerful model may be worthwhile.
One example:
You’re looking for a bike-friendly hotel on Lake Maggiore—preferably affordable, with a restaurant, and for a one-night stay.
Here, the AI must take several requirements into account at the same time:
- Location
- Budget
- Features
- Restaurant
- Bicycle-Friendliness
A model with a medium thinking level takes a little more time, but in return often produces significantly more structured results and takes more details into account.
This extra effort is often worthwhile, especially when conducting research.
Claude is ideal for more complex text-based tasks
In my day-to-day work with Claude, I often use the Sonnet model.
For many tasks, it offers an excellent balance between speed and quality.
A good example is language learning.
If you ask the AI to speak to you in Italian at all times and correct any mistakes, the model has to work much harder than it would for a simple question. It carries on a conversation, keeps track of the context, and provides linguistic feedback at the same time.
A medium thinking level is appropriate for tasks like these.
For extensive research, more powerful models are a good choice
The more complex your query becomes, the more worthwhile it is to use a more powerful language model.
For example:
- Creating Extensive Tables
- Timelines
- historical analyses
- structured comparisons
- extensive research
A prompt such as:
“Create a detailed overview of the development of chatbots since ELIZA. Distinguish between the United States, Europe, and Asia; assess the impact of each development; and include a timeline.”
places significantly greater demands on the model.
In Claude’s case, I’d be more inclined to go with the Opus or a similarly powerful model.
Although these models take a little longer, they often produce more structured and detailed results.
Claude really shines when it comes to Office documents
One of Claude’s strengths is often underestimated.
Especially when creating:
- Word documents
- Excel files
- PowerPoint presentations
Claude often produces very convincing results.
Anyone who regularly creates documents for their daily work should definitely give this feature a try.
ChatGPT is currently my go-to choice for images
When it comes to image generation, ChatGPT is currently my personal favorite.
Image generation works:
- quickly
- reliable
- high-quality
While Claude is very helpful with graphics and charts, ChatGPT is currently much better suited for creative image generation and image editing.
Even more complex image requests are often handled surprisingly well.
So which language model should you choose?
A simple guide might look like this:
| Task | Recommendation |
|---|---|
| Small Talk | Standard Model |
| Translations | Standard Model |
| Simple Questions | Standard Model |
| Search Using Multiple Criteria | Medium model |
| Comprehensive analyses | High-Performance Model |
| Tables and Timelines | High-Performance Model |
| Word, Excel, and PowerPoint files | Claude |
| Create images | ChatGPT |
Conclusion
Don’t let the ever-growing number of language models unsettle you.
For most tasks, a fast standard model is perfectly sufficient. Only when your prompts become more extensive or you want to create complex analyses, research projects, or documents is it worth using more powerful models.
Most importantly, though: Try out the different models for yourselves.
Over time, you’ll develop a sense of which model works best for your specific use cases. It is precisely this practical experience that is often more valuable than any model overview.
Get the guide to the video here: https://www.sophiehundertmark.com/ki-modell-leitfaden
Any further questions?
Do you have any questions? I am happy to support you, act as a sparring partner and answer your questions. I am always happy to receive your messages, preferably by WhatsApp message or e-mail.
