Why does Openai playground GPT4 models disallow images or pdf as input?

[charlie_kilo_romeo]

If I upload an image, or pdf, in playground using any assistant models the response I get is

“I am a AI Text based assistant … and cant view images or files directly”

This may be obvious - from existing posts on this I cant seem to get a clear Yes/No or if this is expected and in that case an ETA on when we could do this on playground (given chatgpt already supports this)

Appreciate help/response.

Rajesh

[konradk]

It’s not possible yet, but they are working on it:

[charlie_kilo_romeo]

cool. I assume it is the case for user messages directly typed in at playground and also the via api (ie, either cannot contain images as of now)

-Raj