MCPcopy Create free account

hub / github.com/BerriAI/litellm / functions

Functions202,479 in github.com/BerriAI/litellm

↓ 5 callersMethodasync_transform_video_content_response
( self, raw_response: httpx.Response, logging_obj: object, client: "AsyncHTTPH
litellm/llms/fal_ai/videos/transformation.py:622
↓ 5 callersFunctionau
({open:e,close:t,initialFocus:r,composer:a,children:s})
litellm/proxy/_experimental/out/_next/static/chunks/2j7hjtph8n74b.js:36
↓ 5 callersMethodaudio_speech
( self, model: str, input: str, voice: str, optional_params: dict,
litellm/llms/azure/azure.py:1211
↓ 5 callersMethodauth
(&self)
litellm-rust/crates/llms/src/base_llm/ocr/handler.rs:76
↓ 5 callersMethodauth_flow
(self, request: httpx2.Request)
litellm/mcp_client/client.py:445
↓ 5 callersFunctionauthorize_model
( identity: &litellm_gateway_auth::AuthenticatedRequest, deployment: &Deployment, body: &Map<Strin
litellm-rust/crates/gateway-inference/src/request.rs:106
↓ 5 callersFunctionautoRouterStrategy
(params: AutoRouterParams | null | undefined)
ui/litellm-dashboard/src/components/add_model/auto_router_strategies.ts:90
↓ 5 callersMethodavalidate_anthropic_messages_environment
( self, headers: dict, # mutable-ok: mirrors the sync validate_anthropic_messages_environment
litellm/llms/anthropic/messages/native/transformation.py:379
↓ 5 callersFunctionavideo_content
Asynchronously download video content. Parameters: - `video_id` (str): The identifier of the video whose content to download - `time
litellm/videos/main.py:395
↓ 5 callersMethodawait_key
(self, key: str, count: int = 1, seconds: float = 30)
tests/integration/_support/anthropic_refusal.py:452
↓ 5 callersFunctionazure_prompt_shield_guardrail_cost
USD cost of an Azure Prompt Shield invocation from its text-record count. Returns 0.0 on the free tier, ``text_records * price / 1000`` when a pr
litellm/litellm_core_utils/llm_cost_calc/guardrail_cost.py:266
↓ 5 callersFunctionazure_spillover
The spillover Azure reports in the response headers, else None.
litellm/litellm_core_utils/ptu_pricing.py:384
↓ 5 callersFunctionb
(e)
litellm/proxy/_experimental/out/_next/static/chunks/0ayyip3c-drm1.js:1
↓ 5 callersFunctionb
({disabledReason:e,onClick:i})
litellm/proxy/_experimental/out/_next/static/chunks/0rgp8d4pvt399.js:1
↓ 5 callersFunctionb
(e,t)
litellm/proxy/_experimental/out/_next/static/chunks/0dvgjd47fte1d.js:1
↓ 5 callersFunctionb
(e,t)
litellm/proxy/_experimental/out/_next/static/chunks/2aatum0yo4cv8.js:1
↓ 5 callersMethodbackfill_team_member_budget_entries
Ensure every team member has a TeamMembership row linked to the team_member_budget. Called after team_member_budget is set/u
litellm/proxy/management_endpoints/team_endpoints.py:962
↓ 5 callersMethodbase
(self)
tests/integration/observability/test_content_filter_logging_only.py:210
↓ 5 callersFunctionbase64_source
(data: JsonValue, media_type: JsonValue = PDF_MEDIA_TYPE)
tests/integration/_support/pdf_document.py:101
↓ 5 callersFunctionbase_counts_cached
`compute` memoized on disk. The base tree at a given commit is immutable, so its counts are a pure function of the merge-base plus the checker's
scripts/lint_base_counts.py:270
↓ 5 callersFunctionbase_url
(gateway: Gateway)
tests/integration/authorization/_responses_compact.py:243
↓ 5 callersMethodbatch_
(self)
litellm/proxy/db/db_spend_update_writer.py:293
↓ 5 callersFunctionbatch_completion_models_all_responses
Send a request to multiple language models concurrently and return a list of responses from all models that respond. Args: *args
litellm/batch_completion/main.py:203
↓ 5 callersMethodbatch_prediction_jobs_handler
Handle batch prediction jobs passthrough logging. Creates a managed object for cost tracking when batch job is successfully created.
litellm/proxy/pass_through_endpoints/llm_provider_handlers/vertex_passthrough_logging_handler.py:854
↓ 5 callersFunctionbd
(s)
litellm/proxy/swagger/swagger-ui-bundle.js:2
↓ 5 callersFunctionbedrock_cache_edge
(store: ResponseStore)
tests/code_coverage_tests/test_provider_cache.py:206
↓ 5 callersFunctionbedrock_edge
(cache: CacheEdge, provider: Provider, action: str = "converse")
tests/code_coverage_tests/test_provider_cache.py:229
↓ 5 callersFunctionbedrock_guardrail_cost_by_unit
USD per counter, keyed like ``usage_units``; None when no pricing entry exists, and None for a counter the entry has no price for, since only an e
litellm/litellm_core_utils/llm_cost_calc/guardrail_cost.py:84
↓ 5 callersFunctionbedrock_llm_proxy_route
Handles Bedrock LLM API calls. Supports both direct Bedrock models and router models from config.yaml. Endpoints: - /model/{modelId
litellm/proxy/pass_through_endpoints/llm_passthrough_endpoints.py:1479
↓ 5 callersFunctionbedrock_native_output_is_opt_in
( model: str, custom_llm_provider: str = "bedrock", model_info: Mapping[str, object] | None = None )
litellm/llms/bedrock/common_utils.py:399
↓ 5 callersFunctionbedrock_peer
(request: Request)
tests/integration/providers/_cache_control_marks_support.py:383
↓ 5 callersFunctionbedrock_request_metadata_headers
The signed ``X-Amzn-Bedrock-Request-Metadata`` header for an Invoke request whose body carries no ``requestMetadata`` of its own. Return
litellm/llms/bedrock/request_metadata.py:227
↓ 5 callersFunctionbedrock_request_metadata_is_owned
Whether the proxy OWNS the request-metadata field and header name for this request. Ownership follows the operator's opt-in alone, never whe
litellm/llms/bedrock/request_metadata.py:190
↓ 5 callersFunctionbilled_usage_cost
The charge an OpenAI-compatible upstream reports in ``usage.cost``, skipped for providers billed from their own usage fields. Once litellm wrote i
litellm/litellm_core_utils/core_helpers.py:1124
↓ 5 callersMethodblocked
(self, key: int = GATE_KEY)
tests/e2e/migrations/database.py:84
↓ 5 callersFunctionblocked_response_usage
Token usage for a synthetic guardrail-blocked response. A post-call block replaces the LLM's response with the violation message, but th
litellm/llms/base_llm/guardrail_translation/utils.py:90
↓ 5 callersFunctionbm25_score_messages
Score each message's relevance to the query using BM25 (Okapi BM25). Parameters: query: The reference text to score against (typical
litellm/compression/scoring/bm25.py:34
↓ 5 callersFunctionbody
(response: httpx.Response)
tests/integration/mcp/test_mcp_user_env_vars.py:35
↓ 5 callersFunctionbounded_http_requests
(gateways: Sequence[Gateway], limit: int)
tests/integration/_support/generation.py:43
↓ 5 callersFunctionbranch_spend_reader
(repository: ConfigRepository, settings: ROISettings)
litellm/proxy/management_endpoints/roi_calculator_endpoints.py:314
↓ 5 callersFunctionbudgetInfoBody
(budgetID string)
terraform/provider/litellm/resource_budget_test.go:12
↓ 5 callersFunctionbuild
()
tests/unit/llms/custom_httpx/test_aiohttp_transport.py:916
↓ 5 callersFunctionbuildTool
(schema: InputSchema | string)
ui/litellm-dashboard/src/app/(dashboard)/mcp-servers/_components/ToolTestPanel.integration.test.tsx:11
↓ 5 callersFunctionbuild_accounting_payload
( payload: SpendLogsPayload, *, entities: Mapping[str, str | None], model_access_groups: Seque
litellm/accounting/snapshot.py:62
↓ 5 callersFunctionbuild_aggregated_sql
(scope: DailyActivityScope, *, api_key_limit: int)
litellm/repositories/daily_activity_sql.py:162
↓ 5 callersFunctionbuild_anthropic_wif_spec
(params: AnthropicWifParams, api_base: str)
litellm/llms/anthropic/wif.py:326
↓ 5 callersFunctionbuild_budget_write_data
Budget-row columns to write. ``budget_reset_at`` tracks any sent ``budget_duration``: recomputed for a new duration, cleared alongside a ``No
litellm/proxy/management_endpoints/organization_endpoints.py:324
↓ 5 callersFunctionbuild_export_sql
( scope: DailyActivityScope, *, export_type: ExportType, after: ExportCursor | None, batch_size: int )
litellm/repositories/daily_activity_sql.py:449
↓ 5 callersMethodbuild_internal_to_public_map
Internal team routing key -> public `team_public_model_name`. Empty when disabled via the legacy flag, the router is absent, or the r
litellm/proxy/common_utils/model_listing_utils.py:428
↓ 5 callersFunctionbuild_jev_request
( prompt: str, system_prompt: str | None, model: str, instructions: str, criteria: Mapping
litellm/router_strategy/complexity_router/jev_classifier.py:442
↓ 5 callersFunctionbuild_kill_switch_request
(config: AgentKillSwitchConfig)
litellm/proxy/agent_endpoints/kill_switch.py:129
↓ 5 callersMethodbuild_mcp_server_table
(self, server: MCPServer)
tests/unit/proxy/mcp_server/test_server_resolution.py:49
↓ 5 callersFunctionbuild_ocr_route
( resources: &litellm_inference::resources::CoreResources, config: &HttpClientConfig, url_policy:
litellm-rust/crates/inference-ocr/tests/support/mod.rs:19
↓ 5 callersFunctionbuild_refinement_prompt
Build a prompt for refining the competitor list based on user instruction.
litellm/proxy/management_endpoints/policy_endpoints/endpoints.py:775
↓ 5 callersMethodbuild_request_log_indexes
()
tests/unit/litellm_proxy_extras/test_litellm_proxy_extras_utils.py:1814
↓ 5 callersFunctionbuild_router_config
The constructor arguments, detached from the caller's objects: settings are static. A bool `cooldown_time` is refused: `True` is not a number of
litellm/router_config.py:84
↓ 5 callersFunctionbuild_string_scrubber
(send_default_pii: bool)
litellm/litellm_core_utils/sentry_scrubbing.py:84
↓ 5 callersFunctionbuild_tool_call_item_kwargs
Build kwargs for an output item dict that is either a ``function_call`` or a ``custom_tool_call`` depending on whether *name* is in *custom_to
litellm/responses/litellm_completion_transformation/custom_tools.py:124
↓ 5 callersMethodbuild_transcription_response
Aggregate a list of normalized "final result" dicts into a ``TranscriptionResponse`` shaped for the requested ``response_format``.
litellm/llms/nvidia_riva/audio_transcription/transformation.py:234
↓ 5 callersFunctionbuild_ui_login_form
( show_deprecation_banner: bool = False, hide_default_credentials_hint: bool = False, )
litellm/proxy/common_utils/html_forms/ui_login.py:14
↓ 5 callersFunctionbuild_user_page_sql
(scope: DailyActivityScope, *, offset: int, limit: int)
litellm/repositories/daily_activity_sql.py:345
↓ 5 callersFunctionbulk_update_keys
Bulk update multiple keys at once. This endpoint allows updating multiple keys in a single request. Each key update is processed ind
litellm/proxy/management_endpoints/key_management_endpoints.py:4311
↓ 5 callersFunctionbulk_update_team_keys
Apply one update payload to many keys inside a single team. Pass `team_id` plus either `key_ids` or `all_keys_in_team=True`. The `update
litellm/proxy/management_endpoints/key_management_endpoints.py:4506
↓ 5 callersFunctionc
(r)
litellm/proxy/_experimental/out/_next/static/chunks/0dvgjd47fte1d.js:1
↓ 5 callersFunctioncK
(e)
litellm/proxy/_experimental/out/_next/static/chunks/1f2vcab0yti03.js:54
↓ 5 callersFunctioncache_access_object
( access_group_id: str, access_group_table: LiteLLM_AccessGroupTable, user_api_key_cache: UserApiK
litellm/proxy/auth/auth_checks.py:3767
↓ 5 callersMethodcache_glob
(self)
scripts/lint_base_counts.py:69
↓ 5 callersMethodcache_key
Identity for processor reuse, so one destination means one exporter. ``span_scope``, the sampling rates, ``capture_message_content`` and
litellm/integrations/otel/model/destination.py:79
↓ 5 callersFunctioncache_request
(input: Value)
litellm-rust/crates/inference/tests/caching.rs:64
↓ 5 callersFunctioncacheable_endpoint
(mount: str, method: str, url: str, body: bytes | None)
tests/e2e/provider_cache.py:176
↓ 5 callersFunctioncacheable_prefix
A prompt prefix above OpenAI's 1024-token caching minimum whose identity is fully determined by `marker` (it is the first word, and prefix caching
tests/e2e/quota_management/spend_tracking/cost_rows.py:210
↓ 5 callersFunctioncall
(method: str, params: dict[str, JsonValue])
tests/integration/compatibility/test_a2a_wire_versions.py:942
↓ 5 callersFunctioncall
(gateway: Gateway, endpoint: Endpoint, group: str, marker: str, priority: int | None = 1)
tests/integration/routing/test_priority_scheduler_queue_cleanup.py:313
↓ 5 callersFunctioncall
(tag: str)
tests/unit/proxy/hooks/test_parallel_request_limiter_v3.py:6142
↓ 5 callersFunctioncall
()
litellm-rust/crates/inference-responses/tests/responses.rs:23
↓ 5 callersFunctioncall
The labels `RedisCache` reports for this operation, in the caller's context so its `service_target` reaches the service event, dispatched on the loop
litellm-rust/crates/python-bridge/src/redis/store/mod.rs:37
↓ 5 callersMethodcall
(self, op: str, request: Mapping[str, object])
litellm/router_core.py:28
↓ 5 callersFunctioncall_config
( py: Python<'_>, kwargs: &Bound<'_, PyDict>, asynchronous: bool, )
litellm-rust/crates/python-bridge/src/http.rs:99
↓ 5 callersFunctioncall_off_loop_with_timeout
Run ``fn`` on a worker thread and await it without blocking the event loop.
litellm/proxy/guardrails/guardrail_hooks/custom_code/bounded_execution.py:174
↓ 5 callersFunctioncall_status
(harness: Harness, video_id: str, *, headers=None, query=None)
tests/unit/proxy/video_endpoints/test_endpoints.py:341
↓ 5 callersFunctioncallback
Actual callback function that Boto3 will call.
litellm/llms/bedrock/common_utils.py:936
↓ 5 callersMethodcallback_list
(self, name: CallbackList)
litellm/router_python_host.py:131
↓ 5 callersFunctioncaller_alias_maps
Team aliases count only when listing the team the key authenticated as.
litellm/proxy/common_utils/model_listing_utils.py:208
↓ 5 callersFunctioncalls_after_a_rate_limit
( core: RouterCore, host: ScriptedHost, **request: object )
tests/unit/test_router/core/test_gaps_request_overrides.py:134
↓ 5 callersMethodcanReplace
(e,t,n=tm.empty,s=0,r=n.childCount)
litellm/proxy/_experimental/out/_next/static/chunks/085r5ieqhajig.js:1
↓ 5 callersMethodcan_rbac_role_call_model
Checks if user is allowed to access the model, based on their role.
litellm/proxy/auth/handle_jwt.py:1452
↓ 5 callersMethodcandidates
The eligible deployments for `name`, in filter-chain order.
litellm-rust/crates/router-core/src/select.rs:259
↓ 5 callersFunctioncanonical_string
(value: str)
tests/e2e/fixture_canonical.py:82
↓ 5 callersMethodcapture
(**kwargs)
tests/unit/integrations/test_rubrik.py:311
↓ 5 callersMethodcapture_sigterm
(cls, clock: Callable[[], float] = time.monotonic)
litellm/proxy/shutdown/graceful_shutdown_manager.py:102
↓ 5 callersFunctioncaptures_span_content
Whether a capture mode puts prompt and response content on spans
litellm/types/utils.py:4047
↓ 5 callersFunctioncarry_superseded_attempt_usage
Carry the superseded stream's partial usage into billing if the request ultimately fails.
litellm/router_utils/fallback_event_handlers.py:462
↓ 5 callersFunctionchange
(label: string, value: string)
ui/litellm-dashboard/src/app/(dashboard)/skills/_components/add_plugin_form.integration.test.tsx:17
↓ 5 callersFunctionchanged_server_tool_call_arguments
( original: Sequence[ChatCompletionAssistantToolCall], rewritten: Sequence[ChatCompletionAssistantToolCall
litellm/llms/anthropic/common_utils.py:2447
↓ 5 callersFunctionchat
(candidate: Gateway, model: str, key: str, marker: str, *, stream: bool = False)
tests/integration/authorization/_guardrail_opt_out.py:53
↓ 5 callersMethodchat
(self, marker: str, *, headers: Sequence[tuple[str, str]] = ())
tests/integration/observability/test_otel_trace_response_header.py:299
↓ 5 callersFunctionchat_count_request
The chat request Responses inference sends a provider with no native Responses API, or None when inference sends this provider the Responses body
litellm/responses/litellm_completion_transformation/count_request.py:173
↓ 5 callersFunctionchat_messages
(marker: str)
tests/integration/routing/test_fallback_hop_foreign_encrypted_reasoning.py:414
↓ 5 callersFunctioncheckTokenValidity
(token: string | null)
ui/litellm-dashboard/src/utils/jwtUtils.ts:25
↓ 5 callersFunctioncheck_api_key_for_custom_headers_or_pass_through_endpoints
( request: Request, route: str, pass_through_endpoints: list[dict] | None, api_key: str, )
litellm/proxy/auth/user_api_key_auth.py:1569
← previousnext →15,501–15,600 of 202,479, ranked by callers