MCPcopy Create free account

hub / github.com/BerriAI/litellm / functions

Functions202,479 in github.com/BerriAI/litellm

↓ 34 callersMethodaclose
(self)
tests/unit/llms/custom_httpx/test_http_handler.py:2086
↓ 34 callersMethodapply_guardrail
( self, inputs: GenericGuardrailAPIInputs, request_data: dict[str, object], in
tests/unit/proxy/guardrails/test_guardrail_endpoints.py:2208
↓ 34 callersMethodasync_log_failure_event
(self, kwargs, response_obj, start_time, end_time)
tests/unit/proxy/hooks/test_prompt_cache_observer.py:204
↓ 34 callersMethodasync_post_call_streaming_iterator_hook
(self, user_api_key_dict, response, request_data)
tests/unit/proxy/utils/proxy_logging/test_guardrail_pipeline.py:2648
↓ 34 callersMethodasync_post_call_success_hook
(self, data, user_api_key_dict, response)
tests/unit/proxy/test_proxy_utils.py:3425
↓ 34 callersMethodasync_pre_call_hook
( self, user_api_key_dict: UserAPIKeyAuth, cache: litellm.DualCache, data: dic
litellm/proxy/guardrails/guardrail_hooks/lakera_ai_v2.py:532
↓ 34 callersMethodasync_set_cache
Asynchronously set cache using run_in_executor to avoid blocking the event loop. Compatible with Python 3.8+.
litellm/caching/s3_cache.py:102
↓ 34 callersFunctionasyncify
Take a blocking function and create an async one that receives the same positional and keyword arguments, and that when called, calls the ori
litellm/litellm_core_utils/asyncify.py:23
↓ 34 callersFunctioncanary
(slot_id: str)
tests/integration/security/_canary.py:117
↓ 34 callersMethodcancel_batch
( self, batch_id: str, *, key: str, provider: str | None = None )
tests/e2e/batches/batch_client.py:221
↓ 34 callersMethodcommit
(self)
litellm/proxy/utils.py:405
↓ 34 callersMethodcreate
()
litellm/proxy/_experimental/out/_next/static/chunks/2_9r2jeem0dh8.js:1
↓ 34 callersFunctioncreate_background_response
( scenario: Scenario, key: str, model: str, *, body: Mapping[str, JsonValue] | None = None
tests/integration/spend/_background_responses.py:166
↓ 34 callersFunctione
(e,t,r)
litellm/proxy/_experimental/out/_next/static/chunks/2aatum0yo4cv8.js:1
↓ 34 callersFunctionflagged
(ids: Sequence[str])
tests/integration/_support/safeguards_rig.py:382
↓ 34 callersMethodfold
(name: &str)
litellm-rust/crates/model-catalog/src/catalog.rs:88
↓ 34 callersFunctionfrom_environment
( environment: Arc<dyn Lookup + Send + Sync>, enterprise_enabled: bool, )
litellm-rust/crates/secrets-cyberark/tests/secret_manager/support.rs:51
↓ 34 callersFunctionfrom_iter
( variables: impl IntoIterator<Item = (OsString, OsString)>, )
litellm-rust/crates/gateway/src/settings.rs:67
↓ 34 callersMethodget_attached_policies
Get list of policy names attached to the given context. Args: context: The request context to match against
litellm/proxy/policy_engine/attachment_registry.py:128
↓ 34 callersMethodget_labels
(label_name: DEFINED_PROMETHEUS_METRICS)
litellm/types/integrations/prometheus.py:984
↓ 34 callersFunctionget_next_standardized_reset_time
Get the next standardized reset time based on the duration. All durations will reset at predictable intervals, aligned from the current time
litellm/litellm_core_utils/duration_parser.py:102
↓ 34 callersFunctionget_request_route
Resolve the request route from the ASGI scope, with ``root_path`` stripped. Prefer this over ``request.url.path`` for any auth, ACL, routing
litellm/proxy/auth/auth_utils.py:1054
↓ 34 callersFunctionhas_attributes
( value: object, shape: type[_Shape], *names: str, )
litellm/litellm_core_utils/core_helpers.py:168
↓ 34 callersMethodjson
(self, *, exclude_none: bool = False)
litellm/proxy/management_endpoints/team_endpoints.py:374
↓ 34 callersMethodload
(path: impl AsRef<Path>)
litellm-rust/crates/config/src/lib.rs:93
↓ 34 callersFunctionmake_kwargs
Build a minimal kwargs dict representative of a litellm callback invocation.
tests/unit/integrations/newrelic/test_newrelic.py:62
↓ 34 callersMethodmodel_dump
(self)
tests/unit/responses/test_streaming_iterator.py:2148
↓ 34 callersFunctiono
()
litellm/proxy/_experimental/out/_next/static/chunks/2_9r2jeem0dh8.js:1
↓ 34 callersFunctionopen_envelope
Validate ``candidate`` and recover the identity and inner grant. Never raises for bad input: every invalid, expired, tampered, or undecryptable
litellm/proxy/mcp_server/outbound_credentials/envelope.py:341
↓ 34 callersFunctionparse_sse
(text: str)
tests/integration/_support/anthropic_sse.py:148
↓ 34 callersFunctionpost
(gateway: Gateway, path: str, body: Mapping[str, JsonValue])
tests/integration/routing/test_priority_scheduler_queue_cleanup.py:287
↓ 34 callersMethodprocess_input_messages
Process input messages by applying guardrails to text content.
litellm/llms/openai/chat/guardrail_translation/handler.py:117
↓ 34 callersMethodread
(self, size: int, /)
litellm/images/utils.py:17
↓ 34 callersMethodredact_text
(&self, text: &str)
litellm-rust/crates/tracing/src/processing.rs:41
↓ 34 callersMethodresolve_agent_access
Agents the key may reach: key and team grants, intersected with the agent's access group ceiling and, for an agent key acting on behalf of an
litellm/proxy/agent_endpoints/auth/agent_permission_handler.py:127
↓ 34 callersMethodset
(self, *args, **kwargs)
tests/unit/test_vcr_redis_persister.py:140
↓ 34 callersMethodsetState
(e)
litellm/proxy/_experimental/out/_next/static/chunks/3lb5gjzjxjxem.js:1
↓ 34 callersMethodset_deployment_health_states
Merge the given states into the shared cache entry, pruning expired ones. Merging instead of replacing lets writers probing different deploym
litellm/router_utils/health_state_cache.py:65
↓ 34 callersMethodsignatures
(self)
tests/integration/authorization/test_rag_team_provider_credentials.py:158
↓ 34 callersMethodstart
(self)
tests/e2e/perf/setup/podcpu.py:105
↓ 34 callersMethodsuperRefine
(e)
litellm/proxy/_experimental/out/_next/static/chunks/2j7hjtph8n74b.js:1
↓ 34 callersMethodtransform_request
( self, model: str, messages: list[AllMessageValues], optional_params: dict,
litellm/llms/oci/chat/transformation.py:637
↓ 34 callersFunctionupstream
(first: Answer, continued: Answer | None = None)
tests/integration/providers/_gemini_long_decode_support.py:144
↓ 33 callersFunctionJ
()
litellm/proxy/_experimental/out/_next/static/chunks/13d402rvlr6kh.js:1
↓ 33 callersFunction_
()
litellm/proxy/swagger/swagger-ui-bundle.js:2
↓ 33 callersFunction_async_chunk_iter
(chunks: List[Any])
tests/unit/proxy/utils/proxy_logging/test_guardrail_pipeline.py:2611
↓ 33 callersFunction_build_iterator
(chunks)
tests/unit/responses/litellm_completion_transformation/test_streaming_iterator_transformation.py:84
↓ 33 callersMethod_call_regular_mcp_tool
Call a regular MCP tool using the MCP client. Args: mcp_server: The MCP server configuration original_tool_n
litellm/proxy/mcp_server/mcp_server_manager.py:6588
↓ 33 callersFunction_check_team_member_model_budget_with_fallback
( valid_token: UserAPIKeyAuth, model_max_budget_limiter: _EveryModelBudgetLimiter, models: Sequenc
litellm/proxy/auth/user_api_key_auth.py:786
↓ 33 callersFunction_deploy
( rig: _Rig, scenario: Scenario, mode: iv.Mode, *, api_key: str = iv.AZURE_KEY, priced: bool = False )
tests/integration/streaming/test_interception_converted_stream_options.py:220
↓ 33 callersFunction_get
(query: str = "")
tests/unit/proxy/management_endpoints/management_v1/test_budgets.py:103
↓ 33 callersFunction_make_agent_mock
(url: str = "http://backend-agent:10001")
tests/unit/proxy/agent_endpoints/test_a2a_endpoints.py:398
↓ 33 callersFunction_make_counter_invalidation_job
Stub spend_counter_store (and user_api_key_cache) so we can observe invalidation calls. Both caches are looked up via ``from litellm.proxy.pr
tests/unit/proxy/common_utils/test_reset_budget_job.py:1390
↓ 33 callersFunction_oauth_error
RFC 6749 section 5.2 / RFC 7591 section 3.2.2 error body. Descriptions carry no token, code, or URL material so they are safe to relay to any clie
litellm/proxy/mcp_server/gateway_dcr_flow.py:439
↓ 33 callersFunction_object_mapping
Return ``value`` as an opaque mapping when it is a dict.
litellm/proxy/guardrails/guardrail_hooks/tool_permission.py:109
↓ 33 callersFunction_passthrough_router
(llm_router: litellm.Router | None)
tests/unit/proxy/pass_through_endpoints/test_passthrough_endpoint_router.py:44
↓ 33 callersFunction_process_group_patch_operations
Process patch operations for a group and return update data, final members and, when the request contained a member ``replace`` op, the absolute t
litellm/proxy/management_endpoints/scim/scim_v2.py:3205
↓ 33 callersMethod_raise_for_provider_error_status
( self, response: httpx.Response, provider_config: Union[ BaseConfig,
litellm/llms/custom_httpx/llm_http_handler.py:6661
↓ 33 callersMethod_row
(model_id, **extra)
tests/unit/proxy/client/cli/test_agents.py:745
↓ 33 callersFunction_team_admin_may_edit
Let team admins change ``fields`` on /team/update for the duration of the block. The registry only lists the fields shipped so far (LIT-5722 adds
tests/unit/proxy/management_endpoints/test_team_endpoints.py:116
↓ 33 callersFunction_v3_request_data
(**overrides)
tests/unit/proxy/guardrails/guardrail_hooks/test_straiker.py:1110
↓ 33 callersMethodanswer
(self, request: Request)
tests/integration/streaming/test_bedrock_guardrail_background_stream_scan.py:48
↓ 33 callersMethodappend
()
litellm/proxy/_experimental/out/_next/static/chunks/242x7dx0zu3hf.js:1
↓ 33 callersMethodapply_guardrail
Moderate prompts (request) and responses (response); fail-open. - ``request``: evaluate the prompt via the before_prompt webhook and
litellm/integrations/rubrik.py:322
↓ 33 callersMethodaretrieve_batch
(self, model: str | None = None, **kwargs: object)
litellm/router.py:280
↓ 33 callersMethodasync_service_success_hook
Arize is used mainly for LLM I/O tracing, sending router+caching metrics adds bloat to arize logs
litellm/integrations/arize/arize.py:229
↓ 33 callersMethodbasename
()
litellm/proxy/_experimental/out/_next/static/chunks/14v3irtjifjt2.js:1
↓ 33 callersFunctioncall_aocr
(server: RecordingServer, **kwargs: object)
tests/test_litellm_rust/ocr/data.py:34
↓ 33 callersFunctioncheck_and_merge_model_level_guardrails
Check if the model has guardrails defined and merge them with existing guardrails in the request data. Args: data: The request data
litellm/proxy/utils.py:9614
↓ 33 callersMethodchunk_processor
( response: httpx.Response, request_body: dict | None, litellm_logging_obj: LiteLLMLog
litellm/proxy/pass_through_endpoints/streaming_handler.py:164
↓ 33 callersFunctioncleanup_file
(client: BatchCleanupClient, file_id: str, *, key: str, provider: str | None = None)
tests/e2e/batches/batch_cleanup.py:57
↓ 33 callersFunctioncreate_pass_through_route
( endpoint, target: str, custom_headers: Mapping[str, object] | None = None, _forward_headers:
litellm/proxy/pass_through_endpoints/pass_through_endpoints.py:2318
↓ 33 callersMethodexecute
(self, statement: LiteralString | sql.Composed, params: tuple[Scalar, ...] = ())
tests/e2e/migrations/database.py:59
↓ 33 callersMethodexpect
(self, *replies: Reply)
tests/integration/_support/provider.py:32
↓ 33 callersFunctionfind_many_in
Rows in chunk order. No take/skip/cursor/order/distinct: none of them survive a split.
litellm/repositories/chunked_in.py:92
↓ 33 callersMethodget_config
(self)
tests/unit/proxy/agent_endpoints/test_endpoints.py:1306
↓ 33 callersMethodhealth
The health view, built once.
litellm/router_rust/core.py:2004
↓ 33 callersFunctionincrement_spend_counters
Atomically increment spend counters for budget enforcement. Uses spend_counter_store (Redis when available, this worker's memory behind it)
litellm/proxy/proxy_server.py:3630
↓ 33 callersMethodjson
(self)
litellm/proxy/guardrails/guardrail_hooks/straiker/straiker.py:181
↓ 33 callersMethodlabels
(self, *labelvalues: object, **labelkwargs: object)
litellm/integrations/prometheus.py:225
↓ 33 callersFunctionmake_runner
( content: bytes = TWO_CHAT_ROWS, concurrency: int = 4, files: Mapping[str, LiteLLM_ManagedFileTab
tests/unit/proxy/batches_endpoints/test_litellm_executed_batches.py:381
↓ 33 callersMethodmap_tool_helper
( self, tool: ChatCompletionToolParam, )
litellm/llms/anthropic/chat/transformation.py:1315
↓ 33 callersMethodmodel_info
Every configured deployment with the price the proxy resolved for it (config override merged over cost-map defaults).
tests/e2e/proxy_client.py:600
↓ 33 callersFunctionregister_plugin
Register a new plugin in the LiteLLM marketplace. LiteLLM acts as a registry/discovery layer. Plugins are hosted on GitHub/GitLab/Bitbuc
litellm/proxy/anthropic_endpoints/claude_code_endpoints/claude_code_marketplace.py:563
↓ 33 callersFunctionregistry_with
(monkeypatch: pytest.MonkeyPatch)
tests/unit/integrations/vector_store_integrations/test_vector_store_pre_call_hook.py:114
↓ 33 callersFunctionresponses_reply
(identity: str, model: str, text: str, *, stream: bool)
tests/integration/_support/openai_wire.py:123
↓ 33 callersMethodretire
(self)
tests/unit/proxy/mcp_server/test_mcp_server_manager.py:19257
↓ 33 callersMethodrun
(self, user_id, server_id, refresh, reread)
tests/unit/proxy/mcp_server/outbound_credentials/test_token_exchanger.py:529
↓ 33 callersFunctionrun_scheduled_ptu_rollup
Run the daily rollup under a cross-pod lock so only one proxy reconciles a day. Every proxy process schedules this cron, and the read-charge-prun
litellm/proxy/spend_tracking/ptu_flat_cost_rollup.py:674
↓ 33 callersMethodstart
(e,t)
litellm/proxy/_experimental/out/_next/static/chunks/2_9r2jeem0dh8.js:1
↓ 33 callersFunctionstart_shadow_eval
Start a shadow eval: duplicate a sampled slice of one or more targets' live traffic against a second arm, judge the two responses blind, and
litellm/proxy/management_endpoints/auto_router_endpoints.py:1714
↓ 33 callersMethodtargets_openai_prompt_cache_breakpoint
( model: str | None, custom_llm_provider: str | None, api_base: object = None,
litellm/integrations/anthropic_cache_control_hook.py:945
↓ 33 callersFunctionto_py
(py: Python<'_>, value: &T)
litellm-rust/crates/host-python/src/marshal.rs:28
↓ 33 callersMethodtranslate_response
Translate an OpenAI ResponsesAPIResponse to AnthropicMessagesResponse.
litellm/llms/anthropic/messages/responses_adapters/transformation.py:739
↓ 33 callersFunctionuK
()
litellm/proxy/_experimental/out/_next/static/chunks/085r5ieqhajig.js:10
↓ 33 callersMethodupdate
( self, where: Mapping[str, object], data: Mapping[str, object],
litellm/proxy/management_endpoints/customer_endpoints.py:97
↓ 32 callersMethodError
()
terraform/provider/litellm/client.go:30
↓ 32 callersFunction_armor_response
(blocked: bool)
tests/unit/proxy/guardrails/guardrail_hooks/test_model_armor.py:2405
↓ 32 callersFunction_attempts
(wire: Wire)
tests/integration/providers/test_bedrock_mantle_output_limit_wire.py:279
↓ 32 callersFunction_chat_body
( model: str, marker: str, *, stream: bool = False, response_format: JsonValue = _RESPONSE
tests/integration/providers/test_vertex_anthropic_structured_output_wire.py:638
← previousnext →1,401–1,500 of 202,479, ranked by callers