fix(openai): preserve reasoning_content for DeepSeek models
by jackjin1997
·
Feb 06, 2026 at 19:28 UTC
·
scan-32be88046719079e
Get this automatically on every PR
Install the Axiomo GitHub App to get Signals as check runs and PR comments on every pull request.
Risk level: Low (20%)
Preserve reasoning_content for DeepSeek models in ChatOpenAI.
openai/langchain_openai/chat_models/base.py line +8 adds logic for reasoning_content
No explicit streaming response handling found
No direct reference to the issue being tested
test_chat_models.py adds 139 new test lines
Has 1 merged PRs to this repo. maintains 64 public repositories. unfamiliar with 4 files being modified.
Review 2 high-priority file(s)
reproduce_34920.py
+54
54 lines changed; New file; Source code
libs/partners/anthropic/tests/unit_tests/test_chat_models.py
+139
139 lines changed; Source code
libs/partners/anthropic/langchain_anthropic/chat_models.py
+34
Source code
libs/partners/openai/langchain_openai/chat_models/base.py
+6
Source code
36
minutes to review
medium
effort level
none
staleness risk
Standard review process
Some concerns to address before approval
Why is security_scan_clean missing? Consider adding this check.
Why is coverage_maintained missing? Consider adding this check.
reproduce_34920.py
New file in high-priority area: 54 lines changed; New file; Source code