Which language models are available in amber?
A language model is the part of amber that understands your question and writes the answer. You can choose which one is used – but you don't have to.
In 3 steps
Open the model menu: In the input field, click the model selector.
If in doubt, choose "Recommended model": amber then decides for itself which model suits your request. That's the right setting for everyday use.
Switch if needed: You can choose and try out a different model at any time – even in the middle of a conversation.
The two categories
You can see which models currently belong to which category in amber under "Usage" → "View price groups". The assignment may change with model updates.
Which one do I use when?
What are tokens?
Language models don't process text as whole words, but in smaller units: tokens. A token can be a whole word, part of a word or a punctuation mark.
How many tokens a request uses depends on four things:
how long your input is
how many documents amber takes into account
how long the answer is
and above all: which model you use.
If you don't have a specific use case and aren't sure which model can help you best, stick with the recommended model.
Good to know
There's no single best model. For most everyday tasks, an Always on model is more than enough. Use Premium deliberately when particularly thorough processing is needed.
Check the sources for important results. Whatever the model – compare with the original document before you pass information on.
Which models are available to you depends on your contract and the admin settings. Individual models may also be temporarily unavailable.