weiqingy opened a new pull request, #906:
URL: https://github.com/apache/flink-agents/pull/906

   Linked issue: #905
   
   ### Purpose of change
   
   The OpenAI Chat Completions chat model defaults to `gpt-3.5-turbo` when the 
caller omits `model`. This is a legacy OpenAI model, superseded by cheaper and 
more capable small models. A user who takes the default silently gets an 
outdated model — and the Java Javadoc example already uses `gpt-4o-mini`, so 
the code base is inconsistent with its own default.
   
   This change updates the default to `gpt-4o-mini` — the current low-cost 
successor to `gpt-3.5-turbo` — in both language runtimes and the docs, keeping 
a sensible zero-config default while dropping the legacy model and keeping Java 
and Python in parity.
   
   - Java — `OpenAICompletionsSetup.DEFAULT_MODEL`
   - Python — `DEFAULT_OPENAI_MODEL`
   - Docs — the `OpenAICompletionsSetup` parameter tables for both languages
   
   ### Tests
   
   No new tests. The existing Python test asserts against the imported 
`DEFAULT_OPENAI_MODEL` constant rather than a hard-coded string, so it 
continues to pin the default without change (`test_openai_chat_model.py`). 
There is no equivalent Java assertion on the default. Verified the Python 
OpenAI chat-model tests pass and the OpenAI Java module compiles.
   
   ### API
   
   No API signature change. Only the value of the default model changes; 
callers that pass `model` explicitly are unaffected.
   
   ### Documentation
   
   - [ ] `doc-needed`
   - [ ] `doc-not-needed`
   - [X] `doc-included`
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to