mirror of
https://github.com/BerriAI/litellm.git
synced 2026-11-10 08:28:25 +00:00
Google AI Studio now answers 404 for gemini-2.5-flash on new projects, so the e2e aliases move to gemini-3.8-flash. The Vertex alias and every Vertex deployment, batch, and passthrough URL use the global location, because gemini-3.8-flash is only served there 3.8 Flash thinks by default, so the Vertex chat tests send reasoning_effort none with more max_tokens, and the Presidio pre_call tests get the same 512 token room as post_call. The custom pricing override rates are scaled to stay about 100x above the new base price |
||
|---|---|---|
| .. | ||
| conftest.py | ||
| jwt_actors.py | ||
| management_client.py | ||
| test_budget_customer_user_org_e2e.py | ||
| test_config_misc_endpoints_e2e.py | ||
| test_jwt_management_e2e.py | ||
| test_key_lifecycle_e2e.py | ||
| test_key_management_e2e.py | ||
| test_management_e2e.py | ||
| test_mcp_lifecycle_e2e.py | ||
| test_model_tag_accessgroup_e2e.py | ||
| test_model_test_connection_e2e.py | ||
| test_team_management_e2e.py | ||