Hi, can we please have the new gpt-4o model in the models list? (On a related note, what do you think of having a free text field in the model drop-down, so one can type the exact model name without waiting for Portkey to update the list?)

Is there a header I can pass in or any other way to make completions/embeddings calls but skip them from Logs?

just bumping the one above, in case it got missed

Also, I am unable to call the https://api.portkey.ai/v1/virtual-keys endpoint. This is what I get:



From the Portkey UI I checked the Key has all the permissions

One of the api calls just returned:

What does this indicate?

Hey team! Just running into this error, could anyone help?

I had one virtual key (say, K1) being used across all my prompts. Then I added two more keys (K2 and K3) and updated all my prompts to use K2. Added a config (C1) to use K2 and K3. Deleted K1. Now when I make an api call:

• without config C1: the response is successful
• with config C1: throws error - portkey_ai.api_resources.exceptions.BadRequestError: {"status":"failure","message":"Following keys are not valid: <k1>"}

This is MacOS + Chrome.

Hi, how do I use the Update Virtual Key endpoint for Bedrock virtual key? The docs suggest it has only one field key, however Bedrock key requires more fields.

Hi folks, any chance is there a terraform provider for portkey?

🐛 The whole playground gets a horizontal scroll bar if for example a codeblock is generated with a long line. I would have expected only the completion box to have a scroll bar.
To reproduce try this prompt: Return a code block with a really long line

Esc to close the Log right sidebar would be nice! : )

🐛 I am unable to save my updated prompt. The API call to https://albus.portkey.ai/v1/prompts/<prompt-uid> gets stuck in a Pending state if my updated prompt has the if conditional variable from mustache.
Example:

🐛 Copy pasting text in to the playground looses original indentation. Sharing more context in the thread:

Hey team, I am unable to open 2 particular logs. As soon as I try open them, I get into an infinite flickering loops of error message. Can someone help me with this @Team Portkey?

Suggesting some UX improvements in the prompt playground, that could have been useful for me:

• One-click button to copy the completion as an assistant message to the left.
• Allow reordering of messages by dragging or some other way

Hey folks! What are some best practices you know that have actually worked to reduce the latency, especially OpenAI?

How can I do more than just variables in Portkey prompts? Exploring more flexibility such as Jinja-like syntax with conditionals, etc.

👋 Noticed Prompt Partials on Portkey, any documentation around the usage and potential use-cases?

Hi team, I have a prompt on Portkey with one system message (and possibly variables). I want to use the prompt completion api to call this prompt but also additionally pass an array of assistant and user messages (think of passing chat history). How can I do this with the portkey api? I am using the python SDK. I looked around the portkey.prompts.completions.create method, but couldn't find anything that could help.

Hey folks, here is a feature request: Slack messages or just webhook events on the back of any changes made to a prompt.

Hi team! 👋

Just wanted to bump here an issue and a PR I opened recently:

https://github.com/Portkey-AI/portkey-python-sdk/issues/76
https://github.com/Portkey-AI/portkey-python-sdk/issues/77

Would be great if this can be included and released! I'm currently working off a fork to unblock myself, but would really prefer to use the original repo directly. 😄