Fix: add OpenRouter Qwen 3.6 Plus metadata

[oss-agent-shin]

Summary

Adds the missing OpenRouter qwen/qwen3.6-plus model metadata to both LiteLLM cost-map files, including context length, output limit, pricing, source URL, and supported capability flags. This fixes local cost-map lookups returning no model info for the new OpenRouter Qwen 3.6 Plus model.

Repro

A local cost-map lookup for openrouter/qwen/qwen3.6-plus previously returned None, so cost/context information was unavailable for that OpenRouter model.

Evidence

Terminal evidence: qwen36_model_metadata_test_evidence.txt

Tests

• Reproduced before fix: pytest tests/test_litellm/test_cost_calculator.py::test_openrouter_qwen36_plus_model_info -q failed with assert None is not None.
• Added tests/test_litellm/test_cost_calculator.py::test_openrouter_qwen36_plus_model_info.
• Passed: pytest tests/test_litellm/test_cost_calculator.py::test_openrouter_qwen36_plus_model_info -q.
• Passed: pytest tests/test_litellm/test_cost_calculator.py -q (39 passed, 12 warnings).
• Passed CI lint scope locally: lock check, frozen sync, Black check, Ruff, MyPy, circular import check, import safety check.

CI

• Available PR checks: all jobs green, including lint, code-quality, validate-model-prices-json, unit-test, provider test groups, proxy test groups, secret-scan, semgrep, and Codecov patch.
• CircleCI pipeline lookup for the fork branch returned no pipeline from the public project API, so there was no CircleCI workflow to triage for this PR.

Review

• No Greptile review or comment was available after the initial CI wait and polling window, so there were no AI-review threads to address.

Relevant issues

Fixes #27424

Linear ticket

N/A

Pre-Submission checklist

• 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?

N/A

CI (LiteLLM team)

• Branch creation CI run
  Link:

• CI run for the last commit
  Link: Available PR checks are green on the latest commit.

• Merge / cherry-pick CI run
  Links:

Screenshots / Proof of Fix

See the terminal evidence and tests above.

Type

🐛 Bug Fix
✅ Test

Changes

• Added openrouter/qwen/qwen3.6-plus to model_prices_and_context_window.json.
• Added the same entry to litellm/model_prices_and_context_window_backup.json.
• Added a regression test for the local cost-map entry.

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}

[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 adds metadata for the openrouter/qwen/qwen3.6-plus model to both cost-map JSON files and includes a regression test to prevent the entry from going missing again. All field values (pricing, context window, output limit, capability flags) were verified against the live OpenRouter listing.

• Adds openrouter/qwen/qwen3.6-plus to model_prices_and_context_window.json and its backup with verified pricing ($0.325/M input, $1.95/M output), 1M token context, 65536 max output, and correct supports_vision, supports_reasoning, and supports_function_calling flags.
• New test in tests/test_litellm/test_cost_calculator.py loads the local cost map and asserts every field, following the established pattern with no network calls.

Confidence Score: 4/5

Safe to merge; the change is additive metadata only and does not touch any request-path code.

The new JSON entry is data-only and all values check out against the OpenRouter listing. The sole issue is that the entry is inserted before qwen3.5-35b-a3b when alphabetical order would place it after — a cosmetic concern with no runtime impact.

No files require special attention; the two JSON files are straightforward metadata additions.

Important Files Changed

Filename	Overview
model_prices_and_context_window.json	Adds openrouter/qwen/qwen3.6-plus entry; pricing ($0.325/M in, $1.95/M out), context window (1M), and max output (65536) all verified against the OpenRouter listing. Entry is inserted slightly out of alphabetical order relative to adjacent qwen3.5 keys.
litellm/model_prices_and_context_window_backup.json	Mirrors the same openrouter/qwen/qwen3.6-plus entry added to the primary JSON; identical values, same ordering concern.
tests/test_litellm/test_cost_calculator.py	New test verifies all fields of the new entry by loading the local cost map (no network calls); consistent with the established pattern in the file.

_{Reviews (1): Last reviewed commit: "Add OpenRouter Qwen 3.6 Plus metadata" | Re-trigger Greptile}

[greptile-apps]

The new qwen3.6-plus entry is inserted immediately before openrouter/qwen/qwen3.5-35b-a3b, placing a higher-numbered model before a lower-numbered one. If this file is maintained in alphabetical key order (as most entries appear to be), the new key should come after qwen3.5-* entries since "3.6" > "3.5" lexicographically. This doesn't affect runtime behavior, but keeping the file sorted makes future diffs and lookups easier.

Suggested change

	"openrouter/qwen/qwen3.6-plus": {
	"input_cost_per_token": 3.25e-07,
	"litellm_provider": "openrouter",
	"max_input_tokens": 1000000,
	"max_output_tokens": 65536,
	"max_tokens": 65536,
	"mode": "chat",
	"output_cost_per_token": 1.95e-06,
	"source": "https://openrouter.ai/qwen/qwen3.6-plus",
	"supports_function_calling": true,
	"supports_reasoning": true,
	"supports_tool_choice": true,
	"supports_vision": true
	},
	"openrouter/qwen/qwen3.5-35b-a3b": {
	"openrouter/qwen/qwen3.5-35b-a3b": {

Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!