docs: add Brainiall LLM Gateway provider guide

docs/tutorials/integrations/llm-providers/brainiall.md

@@ -0,0 +1,110 @@
+---
+sidebar_position: 12
+title: "Brainiall LLM Gateway"
+---
+
+:::warning
+
+This tutorial is a community contribution and is not supported by the Open WebUI team. It serves only as a demonstration on how to customize Open WebUI for your specific use case. Want to contribute? Check out the [contributing tutorial](/contributing).
+
+:::
+
+# Connect Brainiall LLM Gateway
+
+[Brainiall](https://brainiall.com) is an OpenAI-compatible LLM gateway that provides access to **113+ models from 17 providers** through a single API endpoint. It routes requests to AWS Bedrock, giving you access to models from Anthropic, DeepSeek, Meta, Mistral, Amazon, Google, and more — all through the standard OpenAI API format.
+
+## Prerequisites
+
+- **Open WebUI** installed and running
+- **Brainiall API key** — sign up at [brainiall.com](https://brainiall.com)
+
+## Setup via Admin Panel
+
+1. Open **Admin Panel** > **Settings** > **Connections**
+2. Under **OpenAI API**, click **"+ Add Connection"**
+3. Fill in the connection details:
+   - **URL**: `https://apim-ai-apis.azure-api.net/v1`
+   - **API Key**: your Brainiall API key
+4. Click the **checkmark** button to verify and save the connection
+5. Once saved, all available models will auto-populate in the model selector
+
+:::tip
+
+After saving the connection, go to the chat interface and click the model selector dropdown. You should see models prefixed with their provider names (e.g., `claude-opus-4-6`, `deepseek-r1`, `llama-3.3-70b`).
+
+:::
+
+## Setup via Docker Compose
+
+You can also configure the connection through environment variables in your `docker-compose.yml`:
+
+```yaml title="docker-compose.yml"
+services:
+  open-webui:
+    image: ghcr.io/open-webui/open-webui:main
+    container_name: open-webui
+    volumes:
+      - open-webui:/app/backend/data
+    ports:
+      - "3000:8080"
+    environment:
+      - OPENAI_API_BASE_URLS=https://apim-ai-apis.azure-api.net/v1
+      - OPENAI_API_KEYS=your-brainiall-api-key
+    restart: unless-stopped
+
+volumes:
+  open-webui: {}
+```
+
+:::tip
+
+To use Brainiall alongside other OpenAI-compatible providers (such as a local Ollama instance), separate each URL and key with a semicolon:
+
+```
+OPENAI_API_BASE_URLS=https://apim-ai-apis.azure-api.net/v1;http://host.docker.internal:11434/v1
+OPENAI_API_KEYS=your-brainiall-api-key;ollama-key
+```
+
+:::
+
+## Available Models
+
+Brainiall provides access to 113+ models. Here is a selection of popular models available through the gateway:
+
+| Provider | Model | Context Window |
+|----------|-------|----------------|
+| Anthropic | `claude-opus-4-6` | 200K |
+| Anthropic | `claude-sonnet-4-6` | 200K |
+| Anthropic | `claude-haiku-4-5` | 200K |
+| DeepSeek | `deepseek-r1` | 128K |
+| DeepSeek | `deepseek-v3` | 128K |
+| Meta | `llama-3.3-70b` | 128K |
+| Meta | `llama-4-scout-17b` | 128K |
+| Mistral | `mistral-large-3` | 128K |
+| Amazon | `nova-pro` | 300K |
+| Amazon | `nova-micro` | 128K |
+| Qwen | `qwen3-235b` | 128K |
+
+For the full list of available models, use the models endpoint:
+
+```bash
+curl https://apim-ai-apis.azure-api.net/v1/models \
+  -H "Authorization: Bearer your-brainiall-api-key"
+```
+
+## Features
+
+- **OpenAI-compatible API**: works with any tool or framework that supports the OpenAI API format
+- **Streaming support**: real-time token streaming for all chat models
+- **Function calling**: tool use and structured outputs where supported by the underlying model
+- **Single API key**: one key provides access to all 113+ models across 17 providers
+
+## Troubleshooting
+
+- **Connection verification fails**: double-check that the URL ends with `/v1` and that your API key is valid.
+- **No models appear**: click the refresh button next to the connection. If models still do not load, verify your API key has not expired.
+- **Slow responses**: Brainiall routes to AWS Bedrock. Latency depends on the model selected and the size of your request. Smaller models like `nova-micro` or `claude-haiku-4-5` will respond faster.
+
+## Learn More
+
+- [Brainiall website](https://brainiall.com)