LiteLLM

Free plan APISelf-hostedWeb

What LiteLLM is

LiteLLM is an open-source AI gateway and LLM proxy for platform teams. It gives applications a shared OpenAI-compatible API for connecting to more than 140 providers and 1,800 models. Requests can be load balanced, routed to the lowest-cost option, or sent automatically to cheaper models when prompts are simple. Teams can track usage and spending across keys, users, teams, organizations, tools, agents, and MCP, then set hard budgets and rate limits. Fallback routing, virtual API keys, usage analytics, request and response logging, and Prometheus metrics are included in the free Open Source plan.

The gateway can run in a team's infrastructure, including air-gapped environments. Deployment options include an official Helm chart and Terraform module. LiteLLM describes self-hosting as keeping data within that infrastructure without telemetry. Response and semantic caching can use Redis, S3, or GCS. The product also describes hardened non-root images, vulnerability scanning, access controls using SSO, JWT, and RBAC, and secrets-manager integrations. Enterprise adds SSO and SCIM, audit logs, key rotation, organization and team administrators, dedicated support, onboarding, and listed 24/7 support with SLAs. Its pricing FAQ offers an instant 30-day Enterprise trial key without a credit card.

Who it suits

LiteLLM suits platform teams that need one API across multiple LLM providers and want controls for routing, usage, and spend. Enterprise is described for production teams, including high-volume and regulated deployments.

What it does well

One OpenAI-compatible API for 140+ providers and 1,800+ models.
Routes requests by load balance, cost, or prompt simplicity.
Tracks spend across keys, users, teams, and other entities.
Supports hard budgets, rate limits, and fallback routing.
Can be self-hosted, including in air-gapped environments.
Free plan includes virtual keys, logging, and Prometheus metrics.

What to know first

The free Open Source plan is self-hosted.
Enterprise pricing is sized to usage and has no listed price.

Limits

The Open Source plan is self-hosted. Enterprise is billed annually and sized to usage; annual request capacity and pricing depend on deployment architecture and support needs.

Plans

Plan	Billed and limits	Price
Open Source	Free forever · self-hosted · 100+ providers · virtual keys, users and teams · spend tracking · budgets and rate limits · LLM fallbacks · request/response logging · Prometheus metrics	Free
Enterprise	Annual · sized to your usage · Annual request capacity · priced by deployment architecture and support needs · volume discounts · 24/7 support and SLAs	On request

As litellm.ai lists them, read 2 Oct 2026. The maker's own page has the final word.

Getting started

Visit the LiteLLM website.
Choose the free Open Source plan or request the 30-day Enterprise trial key.
For self-hosting, use the official Helm chart or Terraform module.
Connect applications through LiteLLM's OpenAI-compatible API.
Configure routing, budgets, or rate limits as needed.

Verdict

Choose LiteLLM if your platform team needs a shared gateway for many providers, with routing and spend controls. The free plan covers self-hosted gateway features; teams needing Enterprise access controls, audit logs, or dedicated support should consider Enterprise, whose price is sized to usage.

Questions

Is LiteLLM free?

Yes. The Open Source plan costs nothing and is free forever. A 30-day Enterprise trial key is also offered without a credit card.

What does the Open Source plan include?

It includes self-hosting, 100+ providers, virtual keys, users and teams, spend tracking, budgets, rate limits, fallbacks, request and response logging, and Prometheus metrics.

What platforms does LiteLLM support?

Its listed platforms are API, self-hosted, and web.

Can LiteLLM run in a private environment?

Yes. It can be self-hosted, including in air-gapped environments. LiteLLM says self-hosted deployments have no telemetry and keep data within the user's infrastructure.

What does Enterprise add?

Enterprise includes SSO and SCIM, OIDC/JWT authentication, audit logs, secret managers, key rotation, organization and team administrators, dedicated support, and onboarding.

How is Enterprise priced?

Enterprise is billed annually and sized to usage. The annual request capacity and price depend on deployment architecture and support needs.

Facts from LiteLLM's own pages

Caching: Response and semantic caching can use Redis, S3, or GCS. · litellm.ai, 2 Oct 2026
Data handling: LiteLLM says its self-hosted deployment has no telemetry and keeps data within the user's infrastructure. · litellm.ai, 2 Oct 2026
Deployment: The gateway can be self-hosted, including in air-gapped environments, and supports deployment with an official Helm chart or Terraform module. · litellm.ai, 2 Oct 2026
Enterprise access: The Enterprise plan includes SSO and SCIM, OIDC/JWT authentication, audit logs, secret managers, key rotation, and organization and team administrators. · litellm.ai, 2 Oct 2026
Integrations: The integrations documentation lists observability tools including Langfuse, Datadog, Grafana Cloud, OpenTelemetry, LangSmith, Arize/Phoenix, and Helicone. · docs.litellm.ai, 2 Oct 2026
Intended users: The Enterprise plan is described for teams running LiteLLM in production, including high-volume and regulated deployments. · litellm.ai, 2 Oct 2026
Product: LiteLLM is an open-source AI gateway and LLM proxy for platform teams. · litellm.ai, 2 Oct 2026
Routing: LiteLLM supports load balancing, lowest-cost routing, and automatic routing of simple prompts to cheaper models. · litellm.ai, 2 Oct 2026
Security: The product page describes signed hardened non-root images, vulnerability scanning, SSO/JWT/RBAC controls, and secrets-manager integrations. · litellm.ai, 2 Oct 2026
Spend controls: It tracks usage and spend by key, user, team, organization, tool, agent, and MCP, and supports hard budgets and rate limits. · litellm.ai, 2 Oct 2026
Support: Enterprise includes dedicated support and onboarding, with 24/7 support and SLAs listed on the pricing page. · litellm.ai, 2 Oct 2026
Trial: The pricing FAQ offers an instant 30-day Enterprise trial key without a credit card. · litellm.ai, 2 Oct 2026
Unified access: The gateway provides one OpenAI-compatible API to 140+ providers and 1,800+ models. · litellm.ai, 2 Oct 2026