Validate that return_to matches the configured control_plane_url origin. Returns True if return_to is valid and should be used.

Apply cache control injection for the Anthropic-native v1/messages endpoint. Returns (messages, system, remaining_non_message_points).

Async: Send a message to an A2A agent. Uses the @client decorator for LiteLLM logging and tracking. If litellm_params contains custom_ll

Receives - messages: List of dict - messages in the openai format Returns - messages - List[dict] - filtered list of

Runs before the LLM API call to validate and potentially modify input. Uses 'PROMPT' messageType as this is input to the model.

Close the WebSocket

Apply an extractor table to ``source``, dropping ``None`` results.

Derive the deterministic three-character prefix for a server. Uses SHA-256 of ``f"{server_id}#{attempt}"`` and folds the first eight bytes in

fal.ai image generation cost calculator

Calculate the cost of web search requests for X.AI models. X.AI Live Search costs $25 per 1,000 sources used. Each source costs $0.025.

Record a system-wide settings change in LiteLLM_AuditLog. Secret leaves are redacted before the row is written. environment_variables hold ar

Encode provider and model_id into character_id using base64.

Field-level opt-in gate: a non-admin team member may only set `access_group_ids` on a (team) key if their team has opted in by adding

Extract tool names from the request body for the given route. Uses guardrail translation handlers when available, else standalone extractors

Extract TTL from cached messages. Returns the first valid TTL found. Args: messages: List of messages to extract TTL from Retur

Get all of the configured mcp servers for the user in the db with their associated teams ``` curl --location 'http://localhos

Filter and transform beta headers based on provider's mapping configuration. This function: 1. Only allows headers that are present in t

Flush all updates from the queue and return all updates aggregated by entity type.

Aggregate updates by daily_transaction_key.

Extract app roles from the Microsoft Entra ID (Azure AD) id_token JWT. App roles are assigned in the Azure AD Enterprise Application

Must return the base completion call args

Returns a list of deployments that match the requested model and tags in the request. Executes tag based filtering based on the tags in requ

Returns end user object from database or cache. If end user exists but has no budget_id, applies the default budget (if configured via l

For a given model group name, return the combined model info Returns: - ModelGroupInfo if able to construct a model group

Build the AgentCore URL, construct a JSON-RPC envelope, and sign the request. Args: request_id: A2A JSON-RPC request ID

Atomically increment spend counters for budget enforcement. Uses spend_counter_cache (DualCache with Redis backend when available) so co

Retrieve information about a key. Parameters: key: Optional[str] = Query parameter representing the key in the request user_a

Check if a resource ID is a base64 encoded unified ID. Args: resource_id: The resource ID to check prefix: The expected pref

Create GenericGuardrailAPIInputs with tool_calls.

Convert a JSON-RPC response's ``result`` to ``target``. Errors and non-dict results pass through untouched.

Parse tool call arguments from a JSON string. When the JSON is malformed (e.g. truncated by the model), this function attempts a lightwe

Regenerate an existing API key while optionally updating its parameters. Parameters: - key: str (path parameter) - The key to regenerate

Resets the budget for all LiteLLM End-Users (Customers), and Team Members if their budget has expired The corresponding Budget durati

Resolve the auth value for an MCP server. Priority: 1. ``mcp_auth_header`` — per-request/per-user override 2. OAuth2 Token Exchange (OBO

smtp_host, smtp_port, smtp_username, smtp_password, sender_name, sender_email,

Bulk-write all deployment health states as a single cache entry.

Anchor the request's root (server) span for explicit child parenting. No-ops for a non-recordable span so a bad capture can never replace a good

Return the unprefixed name plus the server name used as prefix.

Accept the authorization code from client and exchange it for OAuth token. Supports PKCE flow by forwarding code_verifier to upstream provide

Transform Responses API response to chat completion response

Transform a Responses API request into a Chat Completion request

Get URL and params for retrieving a file.

Transform Search request to Brave Search API format. Transforms Perplexity unified spec parameters: - query → q (same)

Trim a list of messages to fit within a model's token limit. Args: messages: Input messages to be trimmed. Each message is a diction

Update headers with optional anthropic beta.

Async implementation of '_get_cooldown_deployments'

Match production entrypoint: ``_run_deferred_stream_guardrails`` uses dispatch.