BerriAI/litellm - GitHub vs New API
Side-by-side comparison
| Description | LiteLLM is an AI gateway and Python SDK from Berrie AI Incorporated, published in the BerriAI GitHub repository. The SDK provides a common interface for model calls inside Python applications. The proxy gateway centralizes access for a team, with virtual keys, model routing, spend tracking, budgets and an administration interface. The official documentation lists support for more than 100 model providers. Supported endpoints and features vary by integration, so verify your model’s streaming, tool-calling, image, audio or embedding requirements. The router supports retries, fallbacks and load balancing; observability integrations can send request data to tools such as Langfuse, LangSmith and OpenTelemetry. LiteLLM also provides an MCP gateway. It can connect upstream servers using Streamable HTTP, SSE or stdio, expose tools through a fixed gateway endpoint, and scope access by key, team or organization. This requires configuring the upstream servers and authentication; the gateway does not automatically grant access to third-party tools. Agent-to-agent integrations are documented separately. The open-source offering has no software license fee for self-hosting. Code outside the enterprise directory is MIT-licensed, while enterprise code has separate terms. Enterprise pricing is quoted by annual gateway request capacity, deployment architecture and support needs, rather than a per-token license charge. Model-provider charges and infrastructure costs still apply. Enterprise adds controls and support such as SSO, SCIM, audit logs and service-level agreements. Compare New API for another self-hosted gateway with provider-channel management and usage accounting. Evaluate a representative workload, inspect request logging and secret handling, test budget and failure behavior, and decide whether SDK integration or a shared gateway best fits your application. | New API is a self-hosted AI gateway and model-management project maintained in the QuantumNous repository. It brings provider channels, access tokens, model restrictions, usage statistics and cost accounting into one interface. Teams supply their own authorized model-provider access and operate the gateway in their chosen environment. The gateway supports several API formats, including OpenAI-compatible requests, Claude Messages and Google Gemini. Compatibility has boundaries: the README marks Gemini-to-OpenAI conversion as text-only without function calling, and OpenAI-compatible to Responses conversion as in development. Test the exact endpoints, streaming behavior and tool calls your application uses before switching traffic. Routing features include weighted channel selection, automatic retry after failures and user-level model rate limits. The dashboard supports request-based, usage-based and cache-hit cost accounting, with token grouping and model-access controls. These controls can help organize usage, but retries and a common API format do not guarantee uninterrupted service or identical behavior across providers. The current repository license is AGPLv3. Docker Compose is the recommended quick-start path in the README, which also documents Docker commands with SQLite or MySQL. Running the software still involves infrastructure, operations and upstream model charges. Review the current license, secure your deployment and validate accounting against provider bills. Compare LiteLLM if you also want a Python SDK or a documented gateway for MCP servers and agents. |
| Category | Developer Tools | Developer Tools |
| Rating | No reviews | No reviews |
| Pricing | Open Source | Open Source |
| Starting Price | N/A | N/A |
| Use Cases |
|
|
| Tags | AI gatewayPython SDKLLM routingMCP gatewayvirtual keys | AI gatewayLLM routingmodel managementrate limitingusage accounting |
| Features | ||
| Python SDK for direct application integration | ||
| Shared AI proxy gateway and administration UI | ||
| More than 100 documented model-provider integrations | ||
| Virtual keys, users, teams, budgets and rate limits | ||
| Spend tracking and observability integrations | ||
| Router retries, fallbacks and load balancing | ||
| MCP gateway for Streamable HTTP, SSE and stdio upstreams | ||
| Key, team and organization MCP permissions | ||
| Separate enterprise identity, audit and support features | ||
| Self-hosted AI gateway and model management | ||
| Provider channels, token groups and model restrictions | ||
| OpenAI-compatible, Claude Messages and Gemini format support | ||
| Documented limits on protocol conversion | ||
| Weighted channel selection and failure retries | ||
| User-level model rate limiting | ||
| Request, usage and cache-hit cost accounting | ||
| Docker Compose deployment documentation | ||
| View BerriAI/litellm - GitHub | View New API | |