Welcome to Portkey Forum

Updated 10 months ago

Selecting the Appropriate Model

At a glance

The community member is having an issue where only the "sonnet vision" model is showing up when they try to select a model, and they are asking how to resolve this. Other community members respond that the "haiku, sonnet, and opus models are vision models by default", but the "vision part is optional" and the prompts should work even without sending any image data. However, the original community member says it's still not working and they are getting an error saying the "input must be a string" when using the API. They provide a trace ID for one of the calls, and another community member offers to help investigate the issue further if the original community member can share the core request they are sending.

When I try to select the model only sonnet vision shows, maybe thats the issue? How can we resolve it?
V
D
16 comments
@Daniel Yuabov did not get this - all the models are available in the dropdown right?
Can you share what you see?
this is what i see
Attachment
Screenshot_2024-05-08_at_1.39.00_PM.png
We need to see more options for chat
It shows as vision
maybe that is the issue?
haiku, sonnet, and opus models are vision models by default
Attachment
CleanShot_2024-05-08_at_23.56.532x.png
But the vision part is optional. The prompts would work even if you are not sending any image data
yes but its not working
ITs saying input must be a string when I use the API
35367f5f-bf9b-4bfd-bf5a-da422031aaed
THats the trace ID for one of the calls
Thanks for sharing the trace id. I'm trying to see if there's any peculiar issue with your input - because i'm trying out Anthropic sonnet requests and they're working
Would it be possible to share the core request you're sending? Can share in DM
Yes, give me a few minutes and I will send it
Add a reply
Sign up and join the conversation on Discord