Step 4: AI Settings

The language model is the brain and engine of your assistant. Intric is model-agnostic, meaning you can choose from different AI models. Your choice affects performance and where data is processed.

Choose a completion model

Select a completion model (language model). The right model depends on what you want to use it for and your preferences. You can see each model’s speed (how quickly it generates a response) and intelligence (how capable it is at multi-step reasoning and complex tasks).

Note that you may have access to different models in different Spaces depending on the Space’s security class.

Not sure which to choose? Experiment! All models handle a variety of tasks — a more intelligent model is recommended for complex tasks, while a faster model suits simpler ones. We have a dedicated course on language models.

Extra features

Transcription: Lets the assistant handle audio files and convert them to text. When enabled, a microphone icon appears in the chat where you can record audio. Enabling “Include computer audio” lets you share screen audio — ideal for hybrid meetings where both computer and room audio are captured. You can also upload an audio file as an attachment in chat.

Image generation: Lets the assistant create images.

Your assistant is now ready to test!

Test your knowledge

1 questions · 100% correct to pass · Review your answers when done

Start quiz →