8 views
-/https://github.com/berriai/litellm/issues/35187
GitHub · issue

#35187 [Feature]: Allow Playground to Leverage Team-delivered models

  • State: open
  • Author: @cmcfadden
  • Labels: enhancement, ui-dashboard

### Check for existing issues

- [x] I have searched the existing issues and checked that my issue is not a duplicate.

### The Feature

At the moment, the playground is limited to models that a user can access via their personal scope. If a user has no default model access, and only receives model access via Team membership, the playground isn't very useful. They can supply a virtual key for the playground, but the model picker will remain blank.

Instead, the playground should offer a team selector, which updates the model picker and mints an ephemeral key against the team to use for the session, instead of the default UI ephemeral key.

### Motivation, pitch

We're a university and deliver different model sets for different populations, and the teams feature is a great way to do that. However, the playground is also an excellent feature for users, so we want to retain that functionality. For the moment we can work around this issue by just providing all models to users with a $0 budget, but that will be non-viable for compliance reasons as we open up access more broadly.

### What part of LiteLLM is this about?

UI Dashboard

### LiteLLM is hiring a founding backend engineer, …

GitHub resolver

Import GitHub neighbors on demand. Results are saved as system ingests.

Refresh page
vote history (2 events)
#0 of 0 · 31d18h50m6s ago — entered · #import:https:::github.com:berriai:litellm post #2041
The left requires coordinated dashboard, authorization, team-context, and ephemeral-session changes with added state and security considerations. The right is broader across routing and response plumbing, but can likely be implemented as a focused metadata propagation change with targeted compatibility tests.
#0 of 0 · 31d17h57m34s ago — current · #import:https:::github.com:berriai:litellm post #2907
The right-hand task spans dashboard UI, team-scoped model discovery, and secure session credential issuance, requiring coordinated frontend/backend and authorization changes. The left-hand task is a localized provider-auth parameter propagation fix with narrower regression testing.
discussed in #import:https:::github.com:berriai:litellm

ranked child groups

no voted pairs yet in this scope

cli
src
spread
search