Implementation:BerriAI Litellm Supabase Logger

Attribute	Value
Sources	`litellm/integrations/supabase.py`
Domains	Logging, Supabase, Integrations, Observability
Last Updated	2026-02-15 16:00 GMT

Overview

Supabase is a logging integration that records LiteLLM request and response events to a Supabase database table for observability and cost tracking.

Description

The Supabase class connects to a Supabase instance using the SUPABASE_URL and SUPABASE_KEY environment variables. It provides two main methods: input_log_event for recording the initiation of an API call (with status "initiated"), and log_event for recording the final result (with status "success" or "failure"). Each event is stored in a configurable table (default: request_logs) and includes the model name, messages, end user, response time, total cost (computed via litellm.completion_cost), and the litellm_call_id for correlation. The class uses upsert with conflict resolution on litellm_call_id to update existing entries.

Usage

Import and instantiate Supabase when you want to log LiteLLM completion events to a Supabase database. Ensure the environment variables SUPABASE_URL and SUPABASE_KEY are set, and that a request_logs table exists in your Supabase project.

Code Reference

Source Location

litellm/integrations/supabase.py

Signature

class Supabase:
    supabase_table_name = "request_logs"

    def __init__(self)
    def input_log_event(self, model, messages, end_user, litellm_call_id, print_verbose)
    def log_event(self, model, messages, end_user, response_obj, start_time, end_time, litellm_call_id, print_verbose)

Import

from litellm.integrations.supabase import Supabase

I/O Contract

Inputs

Parameter	Type	Description
`model`	`str`	The model name used for the completion call.
`messages`	`list`	The messages sent to the model.
`end_user`	`str`	The end user identifier.
`response_obj`	`dict`	The completion response object (contains `choices` or `error`).
`start_time`	`datetime`	Start time of the API call.
`end_time`	`datetime`	End time of the API call.
`litellm_call_id`	`str`	Unique call identifier for correlation.
`print_verbose`	`Callable`	Verbose logging function.

Outputs

Method	Return Type	Description
`input_log_event`	`None`	Inserts an "initiated" status row; no explicit return value.
`log_event`	`None`	Upserts a "success" or "failure" status row; no explicit return value.

Usage Examples

import os
import litellm

# Set Supabase credentials
os.environ["SUPABASE_URL"] = "https://your-project.supabase.co"
os.environ["SUPABASE_KEY"] = "your-supabase-key"

# Add supabase as a success callback
litellm.success_callback = ["supabase"]
litellm.failure_callback = ["supabase"]

# The Supabase logger is automatically invoked on completion calls
response = litellm.completion(
    model="gpt-3.5-turbo",
    messages=[{"role": "user", "content": "Hello!"}],
)

Related Pages

Implementation:BerriAI_Litellm_S3_Logger - Alternative S3-based logging integration
Implementation:BerriAI_Litellm_Weights_Biases_Logger - Weights & Biases logging integration

Page Connections

Double-click a node to navigate. Hold to expand connections.

Principle

Implementation

Heuristic

Environment