MCPcopy Create free account

hub / github.com/BerriAI/litellm / functions

Functions202,479 in github.com/BerriAI/litellm

↓ 5 callersMethod_no_attribute
What reading an attribute deleted from the core raises (the facade renames it).
litellm/router_rust/core.py:2380
↓ 5 callersFunction_normalize_images_for_message
Ensure each image has an 'index' field, as required by ImageURLListItem. Some providers (e.g. OpenRouter) return images without index.
litellm/litellm_core_utils/llm_response_utils/convert_dict_to_response.py:93
↓ 5 callersFunction_normalize_mcp_server_cost_info
Coerce ``mcp_server_cost_info`` numeric fields to ``float`` at ingest. YAML 1.1 parses scientific notation without a decimal point (e.g. ``7e
litellm/proxy/mcp_server/mcp_server_manager.py:1822
↓ 5 callersMethod_normalize_package_name
Canonicalize package names so '-', '_' and '.' compare equivalently.
tests/code_coverage_tests/check_licenses.py:102
↓ 5 callersMethod_normalize_response_schema
(generate_content_config_dict: dict, model: str)
litellm/llms/gemini/google_genai/transformation.py:312
↓ 5 callersFunction_normalize_responses_caller_body
(body: dict[str, JsonValue])
tests/integration/observability/_openinference_support.py:405
↓ 5 callersFunction_normalize_team_metadata_keys
Coerce a team-metadata allowlist from a list or comma-separated string. config.yaml passes a YAML list; an env var passes a comma-separated strin
litellm/integrations/opentelemetry.py:290
↓ 5 callersFunction_normalized_base
(api_base: str)
litellm/llms/databricks/agent/transformation.py:63
↓ 5 callersFunction_not_found
(request_id: JsonValue, agent: str)
tests/integration/authorization/test_agent_denial_logging.py:119
↓ 5 callersFunction_number
(value: JsonValue)
tests/integration/providers/test_bedrock_mantle_claude_chat_wire.py:190
↓ 5 callersFunction_oauth2_create_payload
(**overrides)
tests/unit/proxy/management_endpoints/test_mcp_management_endpoints.py:7801
↓ 5 callersFunction_oauth_client_table_actions
( prisma_client: PrismaClient, )
litellm/proxy/mcp_server/db.py:598
↓ 5 callersFunction_oauth_error_response
(err: _OAuthError)
litellm/proxy/anthropic_endpoints/gateway_endpoints.py:154
↓ 5 callersFunction_oauth_row
A stored per-user OAuth token row (payload tagged type=oauth2, legacy plain-base64 encoding).
tests/unit/proxy/mcp_server/test_db_credentials.py:219
↓ 5 callersFunction_object_permission_rows
(permission_id: str)
tests/integration/management/test_project_lifecycle.py:97
↓ 5 callersMethod_obo_needs_endpoint_discovery
An ``oauth2_token_exchange`` server with no configured token endpoint can have it discovered (RFC 9728 -> RFC 8414) the same way the ``oauth2`
litellm/proxy/mcp_server/mcp_server_manager.py:2009
↓ 5 callersFunction_obo_protected_resource_response
The OBO (token_exchange) PRM, or None when this server is not OBO / no issuer is configured. The client SSOs with the IdP to obtain a subject tok
litellm/proxy/mcp_server/discoverable_endpoints.py:3107
↓ 5 callersFunction_obo_server
()
tests/unit/proxy/mcp_server/test_mcp_server_manager.py:12192
↓ 5 callersFunction_observation
(path: str, body: Mapping[str, JsonValue])
tests/integration/spend/test_end_user_spend_without_proxy_user.py:73
↓ 5 callersFunction_observations
(gateway: Gateway)
tests/integration/providers/test_vertex_realtime_multi_region_host.py:222
↓ 5 callersFunction_observe
( cache: DualCache, body: Mapping[str, JsonValue], *, deployment_id: str = "sonnet", model
tests/unit/proxy/management_endpoints/test_prompt_cache_prediction.py:91
↓ 5 callersFunction_ocr_logging_kwargs
(call_type: str = "ocr", metadata: _LoggingMetadata | None = None)
tests/unit/proxy/spend_tracking/test_spend_tracking_utils.py:6991
↓ 5 callersFunction_ocr_logging_obj
(litellm_params: dict[str, object])
tests/unit/test_cost_calculator.py:5841
↓ 5 callersFunction_ocr_model
(scenario: Scenario)
tests/integration/compatibility/test_status_code_followups_requests.py:57
↓ 5 callersFunction_of_type
(events: tuple[Event, ...], event_type: str)
tests/integration/providers/test_bedrock_nova_sonic_realtime_tool_calls.py:485
↓ 5 callersFunction_offload_kwargs
()
tests/unit/proxy/hooks/test_proxy_track_cost_callback.py:3315
↓ 5 callersFunction_ollama_function_call_prompt
( messages: list[dict[str, object]], functions: list[object], drop_params: bool = False )
tests/unit/litellm_core_utils/prompt_templates/test_litellm_core_utils_prompt_templates_factory.py:97
↓ 5 callersMethod_on_pidfd_readable
pidfd became readable: engine process exited or became zombie. Sets _engine_confirmed_dead BEFORE cleanup so _run_reconnect_cycle tak
litellm/proxy/utils.py:7338
↓ 5 callersFunction_one
(gateway: Gateway, model: str, call: _Call, *, window: float = _CLIENT_WINDOW)
tests/integration/providers/test_bedrock_stream_timeout_chaos.py:170
↓ 5 callersFunction_one_page_pdf_message
()
tests/unit/proxy/spend_tracking/test_input_tokens.py:36
↓ 5 callersFunction_one_pending_background_response
()
tests/unit/proxy/response_api_endpoints/test_endpoints.py:1118
↓ 5 callersFunction_one_row
(model: str)
tests/integration/routing/test_azure_passthrough_spend_row.py:119
↓ 5 callersFunction_only_request
(wire: Wire)
tests/integration/providers/test_bedrock_runtime_chat_completions_wire.py:125
↓ 5 callersFunction_only_request
(wire: Wire)
tests/integration/providers/test_responses_bridge_tool_search_replay_wire.py:226
↓ 5 callersFunction_only_request
(wire: Wire, provider: Provider, stream: bool)
tests/integration/providers/test_gemini_thinking_replay_wire.py:506
↓ 5 callersFunction_only_upstream
(wire: Wire, target: str)
tests/integration/providers/test_openai_passthrough_multipart_wire.py:109
↓ 5 callersFunction_only_upstream_request
(wire: Wire)
tests/integration/providers/test_anthropic_passthrough_wire.py:129
↓ 5 callersFunction_open_aliases
()
tests/integration/mcp/test_mcp_transports.py:118
↓ 5 callersFunction_open_sealed
Open a sealed value totally: anything that is not prefix-shaped, does not decrypt, or does not validate returns ``None`` for the caller to map ont
litellm/proxy/mcp_server/gateway_dcr_flow.py:497
↓ 5 callersFunction_openai
(gateway: Gateway)
tests/integration/providers/test_api_base_url_join_wire.py:90
↓ 5 callersFunction_openai_api_requests
(openai_api: ControlPlane, key: str)
tests/integration/providers/test_openai_responses_tool_search_wire.py:145
↓ 5 callersFunction_openai_chat_response
(content: str = "scripted reply")
tests/unit/test_router/test_router.py:18326
↓ 5 callersFunction_openai_handler_error
( error_type: str, headers: dict[str, str] | list[tuple[str, str]], status_code: int = 400, me
tests/unit/litellm_core_utils/test_exception_mapping_utils.py:2230
↓ 5 callersFunction_openai_model
(supports_tool_search: bool)
tests/unit/llms/openai/responses/test_openai_responses_transformation.py:114
↓ 5 callersFunction_operation
(**metadata_overrides)
tests/unit/llms/gemini/batches/test_gemini_batches_transformation.py:36
↓ 5 callersFunction_options
(app: str)
tests/unit/proxy/test_dev_reload.py:31
↓ 5 callersFunction_origin_only
Scheme+host for display on the connect page; never the full redirect URI, whose path or query could carry values that do not belong in a page URL
litellm/proxy/mcp_server/gateway_dcr_flow.py:966
↓ 5 callersFunction_otel_span_with_trace_id
(trace_id: int)
tests/unit/proxy/test_litellm_pre_call_utils.py:4659
↓ 5 callersFunction_other_team_row
()
tests/unit/proxy/proxy_server/test_team_model_name_translation.py:312
↓ 5 callersFunction_output_estimator
A `_PROXY_BatchRateLimiter` whose output-token floor is observable: the no-`max_tokens` floor mock returns a distinctive sentinel so tests can
tests/unit/proxy/hooks/test_batch_file_validation.py:1990
↓ 5 callersFunction_output_item_field
(output_item: object, field: str)
litellm/litellm_core_utils/llm_cost_calc/tool_call_cost_tracking.py:34
↓ 5 callersFunction_output_items
Read a response's output items as plain objects; the field is a wide union of item models.
litellm/responses/mcp/mcp_streaming_iterator.py:50
↓ 5 callersMethod_override_auth
(role: LitellmUserRoles)
tests/unit/proxy/ui_crud_endpoints/test_proxy_setting_endpoints.py:3684
↓ 5 callersFunction_own_config
(directory: Path, name: str, section: str, value: JsonValue)
tests/integration/authorization/test_hidden_alias_budget_bypass.py:363
↓ 5 callersFunction_own_team
A team the test deletes itself, so cleanup tolerates the row already being gone.
tests/integration/management/test_team_delete_inputs.py:76
↓ 5 callersFunction_owned_key
(gateway: Gateway, scenario: Scenario, owner: str, **fields: JsonValue)
tests/integration/authorization/test_key_budget_owner_updates.py:66
↓ 5 callersFunction_owned_proxy
(rig: _Rig, directory: Path, setup: _OwnedSetup)
tests/integration/providers/test_malformed_tools_wire.py:1776
↓ 5 callersFunction_owned_record
(created_by, team_id)
tests/unit/enterprise/proxy/hooks/test_managed_files.py:5430
↓ 5 callersFunction_owned_rig
( tmp_path_factory: pytest.TempPathFactory, name: str, line_item_callbacks: JsonValue, *,
tests/integration/observability/test_batch_line_items_cold_storage.py:222
↓ 5 callersFunction_owner_credential
(scenario: Scenario)
tests/integration/authorization/test_key_budget_owner_updates.py:62
↓ 5 callersFunction_pack_config_row
(row: object)
litellm/proxy/utils.py:5495
↓ 5 callersFunction_param_str
(litellm_params: Mapping[str, object] | None, key: str)
litellm/llms/anthropic/wif.py:531
↓ 5 callersFunction_params
(**overrides)
tests/unit/proxy/mcp_server/test_mcp_sampling_completion_flow.py:23
↓ 5 callersFunction_parked_request_ids
(fake_redis: Any)
tests/unit/proxy/utils/prisma_and_spend/test_spend_functions.py:722
↓ 5 callersFunction_parse
(chunk: dict)
tests/unit/llms/gigachat/chat/test_gigachat_chat_streaming.py:12
↓ 5 callersMethod_parse_backend_event
Parse a backend frame once. Returns None for non-JSON or non-object frames.
litellm/litellm_core_utils/realtime_streaming.py:1852
↓ 5 callersFunction_parse_data_url
Parse a data URL to extract the media type and base64 data. Args: data_url: Data URL in format: data:image/jpeg;base64,/9j/4AAQ...
litellm/llms/vertex_ai/gemini_embeddings/batch_embed_content_transformation.py:103
↓ 5 callersMethod_parse_prompt_file
(self, content: str, prompt_id: str)
litellm/integrations/gitlab/gitlab_prompt_manager.py:177
↓ 5 callersFunction_parse_setup
(session_configuration_request: str)
litellm/llms/gemini/realtime/transformation.py:171
↓ 5 callersMethod_parse_sse_data
Parse SSE data line, handling both complete and partial JSON chunks. Args: str_line: The SSE line starting with "data:"
litellm/llms/anthropic/chat/handler.py:1496
↓ 5 callersMethod_parse_token_expiration
Parse the token to extract its expiration time. Returns the datetime when the token expires, or None if parsing fails.
litellm/proxy/db/prisma_client.py:462
↓ 5 callersMethod_parsed_response_body
(raw_response: httpx.Response)
litellm/llms/volcengine/responses/transformation.py:425
↓ 5 callersMethod_partial_jsons
(chunks: list)
tests/unit/llms/anthropic/chat/guardrail_translation/test_anthropic_guardrail_handler.py:401
↓ 5 callersFunction_passthrough_ocr
(result)
tests/unit/llms/azure_ai/passthrough/test_azure_ai_passthrough_transformation.py:401
↓ 5 callersFunction_passthrough_span_name_hook
FastAPI ``server_request_hook``: give passthrough server spans a useful name. The instrumentation matches the route at span creation, so both the
litellm/integrations/otel/mount.py:110
↓ 5 callersFunction_patch_auth
Deprecated no-op kept for call-site compatibility. ``_build_agents_client`` now installs a FastAPI ``dependency_overrides`` entry that inject
tests/unit/proxy/google_endpoints/test_managed_agents_model_param.py:77
↓ 5 callersFunction_patch_base_process
(return_value=None)
tests/unit/proxy/google_endpoints/test_managed_agents_model_param.py:67
↓ 5 callersFunction_patch_client_with_tracker
(manager: MCPServerManager, tracker: _ConcurrencyTracker)
tests/unit/proxy/mcp_server/test_mcp_max_concurrent_requests.py:115
↓ 5 callersFunction_patch_grandfather_env
(monkeypatch, mock_mgr)
tests/unit/proxy/management_helpers/test_object_permission_utils.py:1424
↓ 5 callersFunction_patch_proxy_server_globals
Patch proxy_server's prisma_client and user_api_key_cache to non-None mocks so the helper's None-guard doesn't short-circuit. The actual values do
tests/unit/proxy/auth/test_auth_checks.py:1659
↓ 5 callersFunction_patch_proxy_server_imports
Return a context-manager that stubs _proxy_server_imports so tests don't need a running proxy.
tests/unit/proxy/google_endpoints/test_managed_agents_model_param.py:45
↓ 5 callersFunction_patch_pyroscope_grafana_secrets
Patch proxy_server.get_secret_str for Grafana keys; defer other secrets to the real helper.
tests/unit/proxy/test_pyroscope.py:21
↓ 5 callersMethod_patch_statreload_extra_paths
Make uvicorn's StatReload reloader notice non-Python dev files (the --config YAML and .env). Uvicorn uses WatchFilesReload when the o
litellm/proxy/proxy_cli.py:531
↓ 5 callersFunction_patched_coordination_redis_module_state
Stub every `litellm.proxy.proxy_server` global that `_attach_redis_usage_cache` (and its callers) can write to, shared by the whole coordinati
tests/unit/proxy/test_proxy_server_endpoints_and_startup.py:13206
↓ 5 callersFunction_patched_spend
(value: float)
tests/unit/proxy/auth/test_auth_checks_object_access_and_lookup.py:5631
↓ 5 callersFunction_path
(request: Request)
tests/integration/compatibility/test_openai_sdk_client_paths.py:88
↓ 5 callersFunction_paused_scheduler
(monkeypatch)
tests/unit/proxy/proxy_server/test_proxy_config.py:4781
↓ 5 callersFunction_payload
(**extra)
tests/unit/llms/nadir/test_nadir.py:36
↓ 5 callersFunction_payloads
(text: str)
tests/integration/streaming/test_chat_stream_guardrail_error_events.py:254
↓ 5 callersFunction_pdf_source
(pdf_base64: str)
tests/unit/litellm_core_utils/test_token_counter.py:2719
↓ 5 callersMethod_peek_parked
(self)
litellm/responses/mcp/chat_stream_round_billing.py:229
↓ 5 callersFunction_peer
(reply: bytes)
tests/integration/providers/test_gemini_image_config_and_usage_wire.py:48
↓ 5 callersFunction_peer
(shape: _Shape, gate_after_first: threading.Event | None = None)
tests/integration/sdk/test_sync_stream_connection_release_wire.py:103
↓ 5 callersFunction_pending
(interaction_id: str)
tests/unit/proxy/spend_tracking/test_background_interaction_settlement.py:145
↓ 5 callersFunction_per_message_guardrail_server
Answers one redacted text per chat row it was shown, the way a guardrail that scans per message does, and optionally the rewritten rows themselves
tests/unit/llms/openai/responses/test_openai_responses_guardrail_handler.py:3993
↓ 5 callersFunction_per_message_redactor
()
tests/unit/llms/openai/responses/test_openai_responses_guardrail_handler.py:4013
↓ 5 callersFunction_permissive_proxy_logging
()
tests/unit/proxy/mcp_server/test_mcp_server_manager.py:13587
↓ 5 callersMethod_persist_completed_response_to_cache
(self, *, is_async: bool)
litellm/responses/streaming_iterator.py:844
↓ 5 callersFunction_phoenix_routing_logger
A Phoenix-shaped logger whose owned exporter is a registered factory kind that captures the exporter built per routed header set, so the test can
tests/unit/integrations/otel/test_otel_v2_logger.py:3180
↓ 5 callersFunction_pin_cache
(clock: _ManualClock)
tests/unit/router_utils/test_prompt_caching_cache.py:28
← previousnext →14,701–14,800 of 202,479, ranked by callers