LLM operations

Open-source alternatives to commercial LLM tracing, gateways and evaluation platforms such as LangSmith, Portkey and Arize, self-hosted in your own infrastructure.

LangSmith

Hub for teams moving LLM tracing, evaluation and prompt management off LangSmith, or comparing LangSmith's Enterprise self-hosting with open-source platforms they can run without an enterprise contract.

Portkey

Hub for teams replacing Portkey's hosted or hybrid LLM gateway with a gateway that runs entirely in their own infrastructure, including Portkey's own MIT-licensed gateway.

Arize

Hub for teams moving off Arize AX for LLM tracing and evaluation, including the licensing difference between source-available Phoenix and OSI open-source options.

Braintrust

Braintrust is a commercial AI evaluation and observability platform; self-hosting is Enterprise-only and keeps the control plane with Braintrust. Pages here cover replacing it with self-hosted Langfuse or Opik.

OpenAI API

The OpenAI API serves OpenAI's hosted models per token. Pages here cover self-hosting open-weight models on vLLM or Ollama behind your own gateway when data control, residency, deprecations or cost push teams off the hosted API.

Azure OpenAI

Azure OpenAI, now sold as Foundry Models sold by Azure, runs OpenAI and other models in Microsoft's cloud. Pages here cover replacing it with open-weight models self-hosted on vLLM behind your own gateway.

Amazon Bedrock

Amazon Bedrock is AWS's managed service for foundation models from Amazon and third-party providers. Pages here cover replacing it with open-weight models self-hosted on vLLM and open-source guardrails.

OpenRouter

OpenRouter is a hosted router that gives one API across hundreds of models and providers for a platform fee. Pages here cover replacing it with a self-hosted LiteLLM or Portkey gateway.

Lakera Guard

Lakera Guard, now part of Check Point, is a commercial guardrail API against prompt attacks, harmful content and data leakage. Pages here cover replacing it with self-hosted NeMo Guardrails or Guardrails AI.

Talk to our team.

Tell us what you're working on, whether it's a deployment, an audit, a security test or a cyber range. You'll speak with an engineer who can help you scope it.

  • 30-minute call: free, with no obligation.
  • NDA on request: we can sign before you share details.
  • Clear next steps: a scope and plan after the call.