feat(litellm): record final response cost

[sehr-m]

Summary

• record LiteLLM's final response cost on GenAI spans
• prefer the top-level response_cost value so a supported LiteLLM logging hook can replace the calculated cost before observability callbacks run
• fall back to the standard logging payload for compatibility

This lets applications correct a provider-specific pricing gap without reaching into Sentry's private span state. The consuming getsentry/ops#24340 change uses a LiteLLM CustomLogger logging hook to overwrite response_cost before this integration finishes the span.

Validation

• focused LiteLLM integration test passes in all four static/streaming span configurations on LiteLLM 1.96.0
• Ruff lint and formatting pass
• mypy passes for the full SDK

[github-actions]

Coverage Results 📊

✅ 133666 passed | ⏭️ 7276 skipped | Total: 140942 | Pass Rate: 94.84% | Execution Time: 446m 12s

📊 Comparison with Base Branch

Metric	Change
Total Tests	—
Passed Tests	—
Failed Tests	—
Skipped Tests	—

✨ Test counts unchanged from base.

All tests are passing successfully.

✅ Patch coverage is 100.00% (10 of 10 changed executable lines covered; target 80%).
Project statement coverage is 90.23% (unchanged from base (ab069a5) to head (c97eef5)).

Changed files with executable lines (2)

File	Patch coverage	Changed executable lines
sentry_sdk/consts.py	100.00%	2/2 covered
sentry_sdk/integrations/litellm.py	100.00%	8/8 covered

Coverage diff

@@            Coverage Diff             @@
##        master     #8044       +/-##
==========================================
  Coverage    90.23%    90.23%        —%
==========================================
  Files          202       202         —
  Tracked lines     26660     26669        +9
  Branches      9926      9932        +6
==========================================
+ Hits         24055     24063        +8
- Misses        2605      2606        +1
- Partials      1510      1511        +1

---

Generated by Coverage Action