What is Berri AI
LiteLLM is a gateway designed to provide seamless access to over 100 large language models (LLMs) in the OpenAI format. It offers comprehensive logging, usage tracking, and control features, enabling developers to manage model access, track spending, and set budgets efficiently. Trusted by a large community, LiteLLM ensures high uptime and scalability, making it a reliable solution for enterprises and developers.
How to Use Berri AI
- Logging + Spend Tracking: Log requests, responses, and usage data to platforms like S3, Datadog, OTEL, or Langfuse. Start Here
- Control Model Access: Manage access to models using virtual keys, teams, and model access groups. Start Here
- Budgets & Rate Limits: Set budget limits and track spending across models, keys, teams, and custom tags. Start Here
- Pass-through Endpoints: Migrate projects to the proxy with built-in spend tracking and logging. Start Here
-
OpenAI-Compatible API: Access 100+ LLMs using the OpenAI format across
/chat/completion,/embedding, etc. Start Here - Self-serve Portal: Enable teams to manage their own keys via SSO in a self-serve portal. Start Here
Features of Berri AI
-
Stay in Control
Provides tools for logging, spend tracking, and controlling model access.
-
Logging + Spend Tracking
Logs requests, responses, and usage data to various platforms like S3, Datadog, OTEL, and Langfuse.
-
Control Model Access
Manages access to models using virtual keys, teams, and model access groups.
-
Budgets & Rate Limits
Tracks spending and sets budget limits across models, keys, teams, and custom tags.
-
Move Fast
Offers pass-through endpoints and OpenAI-compatible APIs for seamless integration.
-
Pass-through Endpoints
Enables easy migration of projects to the proxy with built-in spend tracking and logging.
-
OpenAI-Compatible API
Supports access to 100+ LLMs in the OpenAI format across various endpoints.
-
Self-serve Portal
Allows teams to manage their own keys via SSO in a self-serve portal.