From cee9a9e38351ec0c9860ba3d75a19f73c4dc263d Mon Sep 17 00:00:00 2001 From: Seven <246023385+seven7763@users.noreply.github.com> Date: Mon, 3 Aug 2026 10:36:29 +0800 Subject: [PATCH] docs: note OpenAI client base_url for multi-model gateways Document that the OpenAI-compatible api_base_url pattern works with multi-model gateways, using DaoXE (https://api.daoxe.com/v1) as one concrete example. Co-Authored-By: Claude Fable 5 --- .../docs/pipeline-components/generators/openaichatgenerator.mdx | 2 ++ 1 file changed, 2 insertions(+) diff --git a/docs-website/docs/pipeline-components/generators/openaichatgenerator.mdx b/docs-website/docs/pipeline-components/generators/openaichatgenerator.mdx index 2cc3d7ff795..91668a42fe8 100644 --- a/docs-website/docs/pipeline-components/generators/openaichatgenerator.mdx +++ b/docs-website/docs/pipeline-components/generators/openaichatgenerator.mdx @@ -39,6 +39,8 @@ You can pass any chat completion parameters valid for the `openai.ChatCompletion `OpenAIChatGenerator` can support custom deployments of your OpenAI models through the `api_base_url` init parameter. +> **Tip:** The same `api_base_url` pattern works with OpenAI-compatible multi-model gateways — for example [DaoXE](https://daoxe.com/?utm_source=github&utm_medium=organic&utm_campaign=haystack&utm_content=openaichatgenerator) at `https://api.daoxe.com/v1`. + ### Structured Output `OpenAIChatGenerator` supports structured output generation, allowing you to receive responses in a predictable format. You can use Pydantic models or JSON schemas to define the structure of the output through the `response_format` parameter in `generation_kwargs`.