Switching from AssistantsAPI to ResponseAPI - Will there be the equivalent of the Assistants Playground?

[david.captur]

We are investigating the possibility of converting our Assistant API functions to work with the newer Responses API offered by OpenAI. OpenAI have promised to have equivalence of functionality between Assistants and Responses, but will there be the equivalent in the Response API of the Assistants Playground?

Basically, a convenient and simple UI that allows you to control various Assistant traits such as Instructions, Model, Tools (File Search, Code Interpreter), Functions, Model Settings etc:

[beni]

That doesn’t seem to exist yet — but I would love to have that too.

Honestly, I’m even starting to wonder if the way the Assistants Playground works is really the same as the Responses API under the hood.

I’m actually facing another related issue about that here:

Are Responses API and the Assistants API truly equivalent for RAG workflows?

Hello everyone, I’m working on a project using a RAG approach. From my experience, using the same vector_store, model , prompt, instructions, max_num_results, … . I get very different results depending on whether I use: The Assistants playground on the OpenAI website, or The Responses API via the Python SDK In the Assistants playground, the responses are much more accurate and relevant. But with the Responses API, the answers are often vague or even wrong. And the chunks retrieved via fi…