Match litellm.completion supported model parameters with proxy model info

[jyeros]

Relevant issues

Fixes #27717

Linear ticket

Pre-Submission checklist

Please complete all items before asking a LiteLLM maintainer to review your PR

• I have Added testing in the tests/test_litellm/ directory, Adding at least 1 test is a hard requirement - see details
• My PR passes all unit tests on make test-unit
• My PR's scope is as isolated as possible, it only solves 1 specific problem
• I have requested a Greptile review by commenting @greptileai and received a Confidence Score of at least 4/5 before requesting a maintainer review

Delays in PR merge?

If you're seeing a delay in your PR being merged, ping the LiteLLM Team on Slack (#pr-review).

CI (LiteLLM team)

CI status guideline:

• 50-55 passing tests: main is stable with minor issues.
• 45-49 passing tests: acceptable but needs attention
• <= 40 passing tests: unstable; be careful with your merges and assess the risk.

• Branch creation CI run
  Link:

• CI run for the last commit
  Link:

• Merge / cherry-pick CI run
  Links:

Screenshots / Proof of Fix

Type

🐛 Bug Fix

Changes

[jyeros]

@greptileai

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!

[greptile-apps]

Greptile Summary

This PR fixes a mismatch between the model identifier passed to get_optional_params and the actual base model configured in proxy model info. When a proxy model is configured with a base_model in model_info, litellm now uses that base model name to resolve supported optional parameters instead of the proxy alias.

• litellm/main.py: In optional_param_args, the \"model\" key now uses model_info[\"base_model\"] when the field is present and non-empty, falling back to the already-provider-stripped model value otherwise.
• tests/test_litellm/test_main.py: Adds a parametrized mock test that patches get_optional_params and asserts the correct model value is forwarded in both the no-model_info and base_model-present cases.

Confidence Score: 5/5

The change is a single-line conditional in optional_param_args construction; it only affects parameter resolution and does not touch the actual model sent to the upstream provider.

The fix is narrow and well-tested: base_model is only used when model_info is a dict with a non-empty base_model key, with a safe fallback to the original model name in all other cases. The mock test covers both paths without network calls.

No files require special attention.

Important Files Changed

Filename	Overview
litellm/main.py	Uses base_model from model_info (when present) as the model arg to get_optional_params, so proxy model aliases correctly inherit supported parameter sets from their base model.
tests/test_litellm/test_main.py	Adds a parametrized mock-only test verifying that get_optional_params receives the base_model value when model_info contains one, and falls back to the stripped model name otherwise.

_{Reviews (2): Last reviewed commit: "Formatting" | Re-trigger Greptile}

[jyeros]

@greptileai

[oss-pr-review-agent-shin]

🤖 litellm-agent: This PR is currently BLOCKED from merge.

Score: 3/5 ❌

Why blocked:

• 1 PR-related CI failure (triage-llm-invalid: schema validation failed — manual review required) (pr_related_failures, -2 pts)

Details: Score docked for: 1 PR-related CI failure (triage-llm-invalid: schema validation failed — manual review required).

Fix the issues above and push an update — the bot will re-review automatically.

Note: This bot is still in beta and might not always work as expected. Please share any feedback via Slack.