feat(pricing): add groq/qwen/qwen3.6-27b to model cost map

[msdianprince-7]

TLDR

Problem this solves:

• groq/qwen/qwen3.6-27b is served by Groq but missing from the cost map
• cost tracking for it silently returns 0 and get_model_info raises

How it solves it:

• add the model entry with pricing and capability flags from Groq's API
• add a regression test so the values cannot silently drift

Relevant issues

No open issue tracks this; the model is simply absent from the cost map

Linear ticket

Pre-Submission checklist

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

• I have added meaningful tests
• My PR passes all CI/CD checks (e.g., lint, format, unit tests)
• My PR's scope is as isolated as possible; it only solves 1 specific problem
• I have received a Greptile Confidence Score of at least 4/5 before requesting a maintainer review (Greptile reviews automatically once the PR is opened; only comment @greptileai to re-request a review after pushing changes)

Screenshots / Proof of Fix

Captured at commit ca07027 against a live proxy hitting the real Groq API

Before: on any commit without this entry, get_model_info("groq/qwen/qwen3.6-27b") raises "This model isn't mapped yet" and the proxy logs "not in built-in cost map ... cost fields will default to 0", so x-litellm-response-cost comes back 0

After (this commit), config points qwen3.6-27b at groq/qwen/qwen3.6-27b:

curl -sS -D - http://localhost:4001/v1/chat/completions \
  -H "Authorization: Bearer sk-1234" -H "Content-Type: application/json" \
  -d '{"model":"qwen3.6-27b","messages":[{"role":"user","content":"Reply with exactly: pricing works"}],"max_tokens":15}'

HTTP/1.1 200 OK
x-litellm-model-id: 532b52a7c9b7174177e97790238d75e2a45b6bfd6acf5c7aac9bd41cd2f44d14
x-litellm-response-cost: 5.46e-05

usage was 16 prompt + 15 completion tokens, and 16 * 6e-07 + 15 * 3e-06 = 5.46e-05, which matches the header exactly

Type

🆕 New Feature

Changes

Adds groq/qwen/qwen3.6-27b to model_prices_and_context_window.json and the bundled backup, with input 0.60 USD/1M, output 3.00 USD/1M, cached input 0.30 USD/1M, a 131072 context window, 16384 max output tokens, and function calling, tool choice, JSON schema, reasoning and vision flags, all sourced from Groq's models API

Adds tests/test_litellm/llms/groq/test_groq_qwen3_6_27b_metadata.py asserting the resolved model info and the raw cost-map entry so the pricing and capability flags are regression covered

Final Attestation

• The tests check the right things, including the edge cases, and regressions in the respective real-world customer use-cases are not possible after this PR

[CLAassistant]

All committers have signed the CLA.

[greptile-apps]

Greptile Summary

Adds Groq pricing and capability metadata for qwen/qwen3.6-27b to both cost maps, with focused regression coverage for model-info resolution and raw metadata values

Confidence Score: 5/5

The PR appears safe to merge; the new model metadata is consistent across both cost maps and is covered by focused regression tests

The model key, pricing, token limits, and capability flags follow existing Groq cost-map conventions, and the canonical and bundled entries are synchronized

Important Files Changed

Filename	Overview
model_prices_and_context_window.json	Adds the canonical Groq model entry with internally consistent token limits, pricing, and capability fields
litellm/model_prices_and_context_window_backup.json	Mirrors the new model metadata from the canonical cost map without value drift
tests/test_litellm/llms/groq/test_groq_qwen3_6_27b_metadata.py	Adds local, network-free regression tests covering resolved and raw bundled model metadata

_{Reviews (1): Last reviewed commit: "feat(pricing): add groq/qwen/qwen3.6-27b..." | Re-trigger Greptile}

[codspeed-hq]

Merging this PR will not alter performance

✅ 31 untouched benchmarks

---

_{Comparing msdianprince-7:add_groq_qwen3_6_27b_pricing (ca07027) with litellm_internal_staging (b9b27c2)}