Support multi-model comparisons in magic commands

[JasonWeill]

Problem

I want to have the same prompt answered by multiple LLMs with one command.

Proposed Solution

Use LangChain's ModelLaboratory to perform model comparisons by running the same prompt, with the same interpolations, across multiple models.

Additional context

Possible syntax:

%%ai provider1:model1,provider2:model2,alias3,alias4,…
My prompt goes here

[dlqqq]

Users can compare multiple models by using our prompt interpolation feature:

# cell 0
prompt = "what is the square root of pi?"

# cell 1
%%ai <model-id-1>
{prompt}

# cell 2
%%ai <model-id-2>
{prompt}

Since there's been no activity on this issue in the last 18 months, it doesn't seem like this is a feature strongly desired by users. Closing this issue.

To anybody reading this: if you'd like to revive this feature request, please open a new issue and let us know!