Set up LLM providers

Last validated: Sep 29, 2026View as Markdown

Choose a provider guide below to configure its API endpoint, credentials, and models in your Aperture configuration. Compare API formats and requirements in Supported providers and clients before setting up a client.

To forward clients' own provider credentials, follow Set up passthrough mode. You can also configure a shared key as a fallback.

For providers without a dedicated guide, use the OpenAI-compatible provider setup. It configures the /v1/chat/completions API by default. Tailscale does not test or guarantee compatibility with every OpenAI-compatible provider.

Set up passthrough mode

Configure an Aperture provider to forward each client's own credential, including subscription OAuth tokens, to the upstream provider instead of injecting a shared API key.

Set up OpenAI

Configure an OpenAI provider in Aperture so your team can access GPT models.

Set up Anthropic

Configure an Anthropic provider in Aperture so your team can access Claude models.

Set up Google Gemini

Configure a Google Gemini provider in Aperture so your team can access Gemini models using the direct Gemini API.

Set up Gemini Enterprise Agent Platform

Configure a Gemini Enterprise Agent Platform provider in Aperture with a GCP service account and key file so your team can access Gemini and Claude models through Aperture.

Set up Amazon Bedrock

Configure an Amazon Bedrock provider in Aperture so your team can access foundation models through AWS.

Set up Microsoft Foundry

Configure Microsoft Foundry providers in Aperture so your team can access OpenAI GPT and Anthropic Claude models deployed in Foundry.

Set up OpenRouter

Configure an OpenRouter provider in Aperture so your team can access models from multiple providers through a single aggregator.

Set up Vercel AI Gateway

Configure a Vercel AI Gateway provider in Aperture so your team can access models from multiple LLM providers.

Set up a self-hosted provider

Configure a self-hosted or locally running LLM server as a provider in Aperture so your team can access private models through your tailnet.

Set up an OpenAI-compatible provider

Configure an OpenAI-compatible provider in Aperture, including API formats, authorization, and cost estimation.