LiteLLM

AI gateway that puts 100+ LLM providers behind one OpenAI-compatible API with keys and budgets.

Active LLM gateway & observability MIT (core) + LiteLLM Enterprise License (enterprise directory)

Quick answer

LiteLLM is an AI gateway that puts 100+ LLM providers behind one OpenAI-compatible API with keys and budgets, listed under LLM gateway & observability. A typical small-business use: give each team or app its own API key with a monthly budget. Official hardware: minimum 1 vCPU · 4 GB RAM. That fits a Hetzner Cloud CPX22 server at about €19.99/month (server only, excl. VAT). License: MIT (core) + LiteLLM Enterprise License (enterprise directory), which allows internal business use.

Last updated Oct 22, 2026 · facts verified against official docs on Oct 22, 2026 · sources

GitHub stars61.8k

Forks12.6k

Latest releasev1.106.2Oct 21, 2026

Last commitOct 22, 2026

Est. server costfrom €19.99/mo

Activity data comes from GitHub and was last refreshed Oct 22, 2026.

What is LiteLLM?

LiteLLM Proxy gives your apps one endpoint and translates requests to OpenAI, Anthropic, Azure, Bedrock, Vertex, local servers and many more. It adds virtual keys, per-team budgets, rate limits, fallbacks and spend logging. It is also a Python SDK.

Where it fits in a 10–200 person business

Give each team or app its own API key with a monthly budget
Switch model providers without changing application code
Central spend reporting across all AI usage

These use cases are our editorial suggestions, not claims made by the project.

Can a business use LiteLLM? (license)

License	MIT (core) + LiteLLM Enterprise License (enterprise directory)
Type	Open core: Core is MIT/Apache; enterprise folders need a paid license.
Internal business use	Allowed under the license text
What to watch	Everything outside the "enterprise/" directory is MIT. That directory has its own license.

This is a plain-English summary of the license text, not legal advice. Third-party models you run with it have their own licenses.

What hardware does LiteLLM need?

Minimum	1 vCPU · 4 GB RAM
Recommended	Not published

1 vCPU and 4 GiB per Uvicorn worker. 4 GiB is a floor, not a target. On a single VM, set NUM_WORKERS to the vCPU count and scale memory with it.

Source: docs.litellm.ai/docs/proxy/prod

How much does it cost to host LiteLLM per month?

Minimum spec	1 vCPU · 4 GB RAM → CPX22 (2 vCPU, 4 GB, 80 GB disk) €19.99/mo

How we got this: we took the cheapest Hetzner Cloud Regular Performance (shared AMD vCPU) plan with at least the RAM and vCPU the project lists, using prices read from www.hetzner.com/cloud/regular-performance on 2026-10-22. "From" monthly price cap shown by Hetzner, excluding VAT. Prices vary by location. Covers the server only; it excludes LLM API usage, backups, extra storage, GPUs and your time. Disk is not matched; add a volume if you need more than the plan includes.

How to deploy LiteLLM

Method: Docker

docker run -v $(pwd)/litellm_config.yaml:/app/config.yaml \
  -e OPENAI_API_KEY=<your-openai-key> \
  -e LITELLM_MASTER_KEY=sk-<long-random-key> \
  -p 4000:4000 docker.litellm.ai/berriai/litellm:latest \
  --config /app/config.yaml

From the official instructions: docs.litellm.ai/docs/proxy/docker_quick_start. Always check the source for the current version before you run it.

Known gotchas when self-hosting LiteLLM

Without a database, budgets are not enforced. max_budget is silently ignored apart from a one-time startup warning, and virtual keys do not work. source
Memory is not a usable autoscaling signal for the proxy; scale on CPU instead. source

Alternatives in LLM gateway & observability

Langfuse LLM observability with tracing, prompt management, evaluations and cost tracking.

LiteLLM FAQ

How much RAM does LiteLLM need?

The official minimum is 1 vCPU · 4 GB RAM. 1 vCPU and 4 GiB per Uvicorn worker. 4 GiB is a floor, not a target. On a single VM, set NUM_WORKERS to the vCPU count and scale memory with it.

How much does it cost to self-host LiteLLM?

The software is free to download. Based on the official minimum 1 vCPU · 4 GB RAM: That fits a Hetzner Cloud CPX22 server at about €19.99/month (server only, excl. VAT). That covers the server only, not LLM API usage, backups, extra storage or your time.

Is LiteLLM free for commercial use?

LiteLLM is licensed under MIT (core) + LiteLLM Enterprise License (enterprise directory) (Open core). Internal business use is allowed under the license text. Everything outside the "enterprise/" directory is MIT. That directory has its own license. This is a summary, not legal advice.

What are common problems when self-hosting LiteLLM?

Without a database, budgets are not enforced. max_budget is silently ignored apart from a one-time startup warning, and virtual keys do not work. Memory is not a usable autoscaling signal for the proxy; scale on CPU instead.

Is LiteLLM still maintained?

Yes, it is actively developed. Last commit: Oct 22, 2026. Latest release: v1.106.2 (Oct 21, 2026). Checked Oct 22, 2026.

How do I install LiteLLM?

The official method is Docker. The exact commands are in the "How to deploy" section above and in the official instructions: https://docs.litellm.ai/docs/proxy/docker_quick_start

What are self-hosted alternatives to LiteLLM?

Other apps in LLM gateway & observability: Langfuse.

Sources

GitHub repository: github.com/BerriAI/litellm
Official documentation, Hardware requirements, Gotcha 2: docs.litellm.ai/docs/proxy/prod
License text: github.com/BerriAI/litellm/blob/main/LICENSE
Deploy instructions, Gotcha 1: docs.litellm.ai/docs/proxy/docker_quick_start
Hetzner Cloud pricing (cost estimate): www.hetzner.com/cloud/regular-performance
GitHub stars, commits and releases: GitHub REST API, fetched Oct 22, 2026

Facts on this page were checked against the project's own repository and docs on 2026-10-22. Spotted something out of date? Tell us.