From e9dbb49f5d0d19afdcc154f811be9b11a7953f20 Mon Sep 17 00:00:00 2001
From: octo-patch <octo-patch@github.com>
Date: Fri, 22 May 2026 14:59:11 +0800
Subject: [PATCH] docs: add MiniMax as LLM provider via OpenAI-compatible
 endpoint

Add MINIMAX_API_KEY entry to .env.template and a usage section in
README.md so users can run AutoAgent against MiniMax models
(MiniMax-M2.7, MiniMax-M2.7-highspeed) through the existing
openai/* LiteLLM provider with API_BASE_URL=https://api.minimax.io/v1.
---
 .env.template |  1 +
 README.md     | 16 ++++++++++++++++
 2 files changed, 17 insertions(+)

diff --git a/.env.template b/.env.template
index b273695..0df744d 100644
--- a/.env.template
+++ b/.env.template
@@ -9,3 +9,4 @@ GEMINI_API_KEY=
 HUGGINGFACE_API_KEY=
 GROQ_API_KEY=
 XAI_API_KEY=
+MINIMAX_API_KEY=
diff --git a/README.md b/README.md
index 735657a..214a840 100644
--- a/README.md
+++ b/README.md
@@ -394,6 +394,22 @@ DEEPSEEK_API_KEY=your_deepseek_api_key
 COMPLETION_MODEL=deepseek/deepseek-chat auto main
 ```
 
+##### MiniMax
+
+* set the `MINIMAX_API_KEY` in the `.env` file. MiniMax exposes an OpenAI-compatible endpoint, so we route requests through the `openai/` provider in LiteLLM.
+
+```bash
+MINIMAX_API_KEY=your_minimax_api_key
+```
+
+* run the following command to start Auto-Deep-Research.
+
+```bash
+OPENAI_API_KEY=$MINIMAX_API_KEY COMPLETION_MODEL=openai/MiniMax-M2.7 API_BASE_URL=https://api.minimax.io/v1 auto main
+```
+
+> Available models: `MiniMax-M2.7`, `MiniMax-M2.7-highspeed`. Use `https://api.minimaxi.com/v1` as `API_BASE_URL` for the Mainland China region.
+
 
 After the CLI mode is started, you can see the start page of AutoAgent: