MCPcopy Create free account

hub / github.com/BerriAI/litellm / functions

Functions202,479 in github.com/BerriAI/litellm

↓ 5 callersFunctionremove_additional_properties
clean out 'additionalProperties = False'. Causes vertexai/gemini OpenAI API Schema errors - https://github.com/langchain-ai/langchainjs/issues/52
litellm/utils.py:5141
↓ 5 callersFunctionremove_sensitive_info_from_deployment
Removes sensitive information from a deployment dictionary. Args: deployment_dict (dict): The deployment dictionary to remove sensit
litellm/proxy/common_utils/openai_endpoint_utils.py:18
↓ 5 callersFunctionremove_strict_from_schema
Relevant Issues: https://github.com/BerriAI/litellm/issues/6136, https://github.com/BerriAI/litellm/issues/6088
litellm/utils.py:5167
↓ 5 callersFunctionrender
(case: PerformanceTestCase, observation: Observation, rows: list[Row])
tests/e2e/perf/setup/observe.py:622
↓ 5 callersMethodrender
()
litellm/proxy/_experimental/out/_next/static/chunks/1-f3yvj3y6lg8.js:5
↓ 5 callersFunctionrender_subtitle_tokens_as_vtt
Render tokens as a WebVTT document; the WEBVTT header is emitted even without cues.
litellm/litellm_core_utils/audio_utils/subtitle_utils.py:230
↓ 5 callersFunctionreplaceAt
(items: readonly T[], index: number, value: T)
ui/litellm-dashboard/src/app/(dashboard)/playground/components/systemOneUI/lib/formPayload.ts:58
↓ 5 callersFunctionreplacement
(name: str, **fields: JsonValue)
tests/integration/management/test_claude_code_plugin_installation_preference.py:49
↓ 5 callersFunctionreplay_stream
(root: Path)
tests/e2e_harness/test_provider_edge.py:984
↓ 5 callersMethodreply
(self, key: str, model: str, prompt: str, *, with_tool: bool = False)
tests/e2e/llm_translation/conversational_matrix.py:308
↓ 5 callersMethodreport
(self)
tests/e2e/perf/test_redis_chaos_e2e.py:139
↓ 5 callersFunctionrequest
()
tests/unit/proxy/test_route_llm_request.py:2738
↓ 5 callersFunctionrequest_litellm_call_id
(data: Mapping[str, object])
litellm/proxy/common_request_processing.py:2143
↓ 5 callersFunctionrequest_root_http_route
``http.route`` exactly as the request's root SERVER span reports it. Read off the span rather than re-derived, so the LLM call span cannot disagr
litellm/integrations/otel/plumbing/context.py:84
↓ 5 callersMethodrequested
(self, kwargs: Mapping[str, object])
litellm/rust_bridge/public_call.py:90
↓ 5 callersFunctionrequire_harness
Return harness if it is a Harness member, else raise TypeError with a hint.
litellm/harness/types.py:34
↓ 5 callersFunctionresample_to_riva_pcm
Decode ``file_bytes`` and produce 16 kHz mono LINEAR_PCM (int16 little endian) suitable for streaming to Riva, plus the audio duration in
litellm/llms/nvidia_riva/audio_transcription/audio_utils.py:64
↓ 5 callersFunctionresetToNonEmailMatchState
()
litellm/proxy/swagger/swagger-ui-bundle.js:2
↓ 5 callersFunctionresolve
(model: string)
ui/litellm-dashboard/src/lib/autorouter_presets.ts:318
↓ 5 callersFunctionresolve
( verified: VerifiedIdentity, identities: &dyn IdentityResolver, authorizer: Arc<dyn Authorizer>,
litellm-rust/crates/gateway-auth/src/authentication.rs:154
↓ 5 callersMethodresolve
( self, server_id: str, caller: UserAPIKeyAuth, *, is_admin_view: bool
litellm/proxy/mcp_server/server_resolution.py:147
↓ 5 callersMethodresolve
( &mut self, py: Python<'_>, arguments: &Bound<'_, PyDict>, )
litellm-rust/crates/python-bridge/src/routes/mod.rs:74
↓ 5 callersMethodresolve
The environment is only for what Python reads straight from `os.environ`, `REDIS_USERNAME` for Azure.
litellm-rust/crates/gateway/src/redis/configured.rs:124
↓ 5 callersMethodresolve_admitted_subject_servers
Union of what each of the admitted subject's sources reaches, each answered by the canonical resolver so no rule is reimplemented for this cal
litellm/proxy/mcp_server/auth/user_api_key_auth_mcp.py:2489
↓ 5 callersMethodresolve_agent_config_from_registry
Resolve agent configuration from the registry for a registered agent. Args: agent_name: The model string with the provid
litellm/llms/a2a/chat/transformation.py:73
↓ 5 callersMethodresolve_anthropic_api_base
(api_base: str | None = None)
litellm/llms/anthropic/common_utils.py:1665
↓ 5 callersFunctionresolve_auth
( services: &AuthServices, validated: ValidatedEnvironment, env_lookup: &(dyn Fn(&str) -> Option<S
litellm-rust/crates/llms/src/base_llm/auth.rs:56
↓ 5 callersFunctionresolve_base_point
The snapshot commit base counts are measured at: merge-base(base_ref, HEAD), made aware of an in-progress merge. Mid-merge, HEAD is still the pre-
scripts/lint_base_counts.py:105
↓ 5 callersFunctionresolve_batch_enqueued_token_scopes
( user_api_key_dict: UserAPIKeyAuth, )
litellm/proxy/hooks/batch_enqueued_tokens.py:135
↓ 5 callersFunctionresolve_credentials_with_source
( provider: UpstreamCredentialProvider, subject: Subject, server: ServerSpec )
litellm/proxy/mcp_server/outbound_credentials/resolver.py:522
↓ 5 callersFunctionresolve_default_end_user_budget
The default budget for an end user with no budget of its own. The key's ``end_user_budget_id`` takes precedence over the proxy-wide ``li
litellm/proxy/auth/auth_checks.py:1885
↓ 5 callersFunctionresolve_deployment
( gateway: &'a Gateway, route: &'static str, body: &Map<String, Value>, )
litellm-rust/crates/gateway-inference/src/request.rs:126
↓ 5 callersFunctionresolve_embedding_max_input_tokens
Explicit cache setting first, else the Router deployment's configured ``max_input_tokens``.
litellm/caching/_embedding_router.py:50
↓ 5 callersFunctionresolve_llm_provider
( model: &'a str, custom_llm_provider: Option<&'a str>, route: &'static str, )
litellm-rust/crates/inference/src/provider.rs:13
↓ 5 callersFunctionresolve_local_oidc_reference
Only local kinds resolve, since request paths must not do network I/O, and ``oidc/env_path/`` is excluded because it bypasses the ``oidc/file/`` c
litellm/secret_managers/main.py:212
↓ 5 callersFunctionresolve_mappers
Resolve mapper names to instances. Unknown names raise ``ValueError``.
litellm/integrations/otel/mappers/__init__.py:41
↓ 5 callersMethodresolve_oauth2_flow_for_request
Return ``server`` with its effective oauth2_flow applied, for egress paths. A stamped row is returned unchanged (its effective flow equals th
litellm/proxy/mcp_server/mcp_server_manager.py:1984
↓ 5 callersFunctionresolve_openai_bearer_token_async
( static_api_key: str | None, api_base: str | None, litellm_params: Mapping[str, object] | None, )
litellm/llms/openai/workload_identity.py:127
↓ 5 callersFunctionresolve_plugin_version
( requested: str | None, current: str | None = None, *, updating: bool = False )
litellm/repositories/plugin_publication.py:60
↓ 5 callersFunctionresolve_provider
Map a litellm provider string to a ``gen_ai.provider.name`` value. Unknown providers pass through verbatim — the convention explicitly allows
litellm/integrations/otel/model/semconv.py:474
↓ 5 callersFunctionresolve_s3_max_retry_age_seconds
(configured: object, fallback: int | None)
litellm/integrations/s3.py:81
↓ 5 callersFunctionresolve_sso_config
Resolve the effective SSO config: stored row first, then process env. Decryption happens here, once, via the pure ``decrypt_value_helper``; this
litellm/proxy/config_resolvers/sso.py:81
↓ 5 callersFunctionresolve_trace_read_scope
( auth: UserAPIKeyAuth, permitted_team_lookup: Callable[[], Awaitable[Sequence[str]]], )
litellm/proxy/auth/authorization.py:98
↓ 5 callersMethodresolved
(self)
litellm/router_rust/core.py:1348
↓ 5 callersFunctionresolved_token_header
Which upstream header the value ``resolve_mcp_auth`` just returned belongs in. ``None`` means keep the auth_type default. A caller-supplied ``mcp
litellm/proxy/mcp_server/oauth2_token_cache.py:350
↓ 5 callersFunctionresourceLiteLLMBudgetRead
(d *schema.ResourceData, m interface{})
terraform/provider/litellm/resource_budget.go:184
↓ 5 callersFunctionresourceLiteLLMJWTKeyMappingUpdate
(d *schema.ResourceData, m interface{})
terraform/provider/litellm/resource_jwt_key_mapping_crud.go:95
↓ 5 callersFunctionresourceLiteLLMKeyBlockRead
(d *schema.ResourceData, m interface{})
terraform/provider/litellm/resource_key_block.go:77
↓ 5 callersFunctionresourceLiteLLMModelCreate
(d *schema.ResourceData, m interface{})
terraform/provider/litellm/resource_model_crud.go:373
↓ 5 callersFunctionresourceLiteLLMPromptRead
(d *schema.ResourceData, m interface{})
terraform/provider/litellm/resource_prompt.go:200
↓ 5 callersFunctionresourceLiteLLMTagRead
(d *schema.ResourceData, m interface{})
terraform/provider/litellm/resource_tag.go:197
↓ 5 callersFunctionresourceLiteLLMTeamMemberAdd
()
terraform/provider/litellm/resource_team_member_add.go:11
↓ 5 callersFunctionresources
()
litellm-rust/crates/python-bridge/src/http.rs:89
↓ 5 callersMethodrespond
(self, request: httpx.Request)
tests/unit/integrations/generic_api/test_generic_api_callback.py:726
↓ 5 callersFunctionresponse
(id: &str, output: Value)
litellm-rust/crates/inference-responses/src/ids.rs:295
↓ 5 callersFunctionresponse_format_schema_from_request
(data: str, json_mode: bool | None)
litellm/llms/bedrock/chat/invoke_handler.py:189
↓ 5 callersFunctionresponse_identities
(value: str)
tests/integration/_support/responses_vendor.py:98
↓ 5 callersFunctionresponse_pattern
(rows: tuple[SpendLogRecord, ...])
scripts/seed_tracing_fixtures.py:122
↓ 5 callersFunctionresponses_binding
(native: NativeResponses | None)
tests/test_litellm_rust/responses/test_dispatch.py:39
↓ 5 callersFunctionresponses_continuation_input
The prior input, the partial response as assistant history, and a user request to continue it.
litellm/router_utils/pre_call_checks/continuation_prefill_check.py:334
↓ 5 callersFunctionresponses_continuation_input_for
The continuation input as the deployment receives it, without the request to continue where the deployment takes the partial response as prefill.
litellm/router_utils/pre_call_checks/continuation_prefill_check.py:344
↓ 5 callersMethodresponses_tools_to_chat_forms
(tools: ResponseTools)
litellm/responses/litellm_completion_transformation/transformation.py:3017
↓ 5 callersMethodrestore_transactions_to_redis
( self, daily_tag_spend_update_transactions: dict[str, DailyTagSpendTransaction] )
tests/unit/proxy/db/test_db_spend_update_writer.py:5410
↓ 5 callersFunctionrestrictedBy
( value: { custom_tier_set?: CustomTierSet; classifier_type?: string }, key: keyof typeof CUSTOM_TIER_REST
ui/litellm-dashboard/src/components/add_model/TierRestrictions.tsx:6
↓ 5 callersFunctionresult_properties
The custom signals a standard reporter cannot derive. Loki, Grafana and tests/integration/conftest.py read the `package`/`covers`/`source` prefix
tests/e2e/junit_properties.py:91
↓ 5 callersMethodresume
( self, thread_id: str, prompt: str, *, model: str, cwd: Path, compact_after_tokens: int | None = None
tests/e2e/codex/harness.py:98
↓ 5 callersMethodresume
(&self)
litellm-rust/crates/redis-testing/src/cluster.rs:149
↓ 5 callersMethodresume
( &mut self, py: Python<'_>, interruption: Option<HostFailure<<M::Protocol as Protocol
litellm-rust/crates/host-python/src/native.rs:62
↓ 5 callersFunctionresumed
(saved_in: str)
tests/unit/llms/opencode/harness/test_transformation.py:636
↓ 5 callersFunctionresumed_session_arg
(sandbox: FakeSandbox)
tests/unit/llms/opencode/harness/test_transformation.py:645
↓ 5 callersFunctionretry_attempts
(default: int)
tests/e2e/e2e_http.py:296
↓ 5 callersMethodreturn_processed_chunk_logic
( # noqa: C901 self, completion_obj: dict[str, Any], model_response: ModelResponseStr
litellm/litellm_core_utils/streaming_handler.py:1259
↓ 5 callersFunctionrevoke_credential
( revocation_endpoint: str, client_id: str, refresh_token: str, http: Http )
litellm/proxy/client/cli/commands/pkce_login.py:405
↓ 5 callersFunctionrewrite
(assembled: ModelResponse)
tests/unit/proxy/guardrails/test_anthropic_sse.py:40
↓ 5 callersFunctionrewrite_request_model
Point the auth-time payload, the parsed-body cache, ``request.json()`` and ``request.body()`` at ``model``. The cache and raw body keep only the k
litellm/proxy/common_utils/http_parsing_utils.py:483
↓ 5 callersFunctionrewritten_item
(client_event: Mapping[str, object], texts: tuple[str, ...])
litellm/litellm_core_utils/realtime_guardrails.py:113
↓ 5 callersFunctionrollUpBreakdown
( results: readonly DailyData[], dimension: BreakdownDimension | "api_keys", )
ui/litellm-dashboard/src/app/(dashboard)/usage/_components/components/overview/overviewData.ts:24
↓ 5 callersMethodrotate_guardrail_params_master_key
Re-encrypt every guardrail row's sensitive litellm_params under the key the proxy decrypts with after the rotation (LITELLM_SALT_KEY when set,
litellm/proxy/guardrails/guardrail_registry.py:558
↓ 5 callersFunctionrotate_mcp_server_credentials_master_key
(prisma_client: PrismaClient, touched_by: str, new_master_key: str)
litellm/proxy/mcp_server/db.py:1538
↓ 5 callersFunctionrouteOf
(item: MenuItem)
ui/litellm-dashboard/src/components/leftnav.tsx:519
↓ 5 callersFunctionroute_error_to_pyerr
(error: Error)
litellm-rust/crates/python-bridge/src/errors.rs:147
↓ 5 callersMethodroute_streaming_logging_to_handler
Route the logging for the collected chunks to the appropriate handler Supported endpoint types: - Anthropic - Vertex
litellm/proxy/pass_through_endpoints/streaming_handler.py:329
↓ 5 callersFunctionrouter
( route_config: SwitchyardRouterConfig, *, completion=unexpected_judge, decide=unexpected_deci
tests/unit/router_strategy/switchyard_router/test_switchyard_router.py:48
↓ 5 callersFunctionrouter_host
The installed host. Nothing installed: the proxy's host if the proxy imports, else the default.
litellm/router_host.py:99
↓ 5 callersFunctionrouter_settings
(cache: OwnedRedis, **extra: JsonValue)
tests/integration/routing/test_payment_required_cooldown.py:113
↓ 5 callersMethodrouting_callbacks
``_routing_callbacks``: the host's callbacks, or after ``observe(request_scoped=True)`` the host's callbacks with the router's own selectors a
litellm/router_rust/steps/callbacks.py:442
↓ 5 callersMethodrouting_group
The routing group callable as ``name``, or None (was ``get_routing_group``).
litellm/router_api.py:1039
↓ 5 callersFunctionrpc
(gateway: Gateway, method: str, params: dict[str, JsonValue], *, status: int = 200)
tests/integration/mcp/test_interactions.py:206
↓ 5 callersFunctionrt
litellm/proxy/_experimental/out/_next/static/chunks/2aatum0yo4cv8.js:1
↓ 5 callersFunctionrun
A failure renders each mode's own quiet fallback: Claude Code gets the label it already knows, Codex gets nothing at all rather than a bare string
litellm/proxy/client/cli/commands/statusline_script.py:426
↓ 5 callersMethodrun
(self, fetch: Callable[[], _R], while_busy: _R | None = None)
litellm/litellm_core_utils/single_flight.py:19
↓ 5 callersMethodrun
( self, user_id: str, server_id: str, refresh: Callable[[], Awaitable[OAuthTok
litellm/proxy/mcp_server/outbound_credentials/oauth_token_store.py:206
↓ 5 callersMethodrun_all
(self)
tests/unit/llms/base_llm/auth/test_token_exchange.py:122
↓ 5 callersFunctionrun_id_from_sse_frames
(frames: bytes)
litellm/proxy/pass_through_endpoints/llm_provider_handlers/tinyfish_passthrough_logging_handler.py:84
↓ 5 callersFunctionrun_migrations
( database: &ClickHouseDatabase, migrator: &Migrator, schema: &str, render: R, )
litellm-rust/crates/storage-clickhouse/tests/migrations.rs:88
↓ 5 callersFunctionrun_model_listing
(listing: Callable[[], _Listed])
litellm/proxy/auth/model_checks.py:56
↓ 5 callersFunctionrun_spend_capture_rate_check
Check every configured provider, publish each rate, and alert on every outcome that warrants one.
litellm/proxy/spend_tracking/spend_capture_rate.py:248
↓ 5 callersFunctionrun_through
(host: &RecordingHost)
litellm-rust/crates/inference-messages/tests/messages/host.rs:246
↓ 5 callersFunctionrust_ocr
(request: pytest.FixtureRequest)
tests/unit/rust_bridge/ocr/test_secrets.py:105
↓ 5 callersFunctions
(e)
litellm/proxy/_experimental/out/_next/static/chunks/2s7apn3nfa3q3.js:1
← previousnext →16,501–16,600 of 202,479, ranked by callers