# Introduction Private AI for financial services: open-weight models behind APIs that work with OpenAI and Anthropic clients, and a full chat app. PrivateMind runs open-weight models for your organisation, on GPUs pooled across customers or dedicated to you. Your data never leaves your environment and is never used to train a model. We built it for financial services, where compliance, data sovereignty and auditability come first. - [Using the API](https://docs.privatemind.com/quickstart.html): Keep your OpenAI or Anthropic SDK and change the base URL. - [Using the app](https://docs.privatemind.com/chat.html): Chat, files, projects, automations and code execution in your browser. - [API Explorer](https://docs.privatemind.com/api-explorer.html): Browse every endpoint and send live requests from the browser. - [Deploy a model](https://docs.privatemind.com/model-deployments.html): Serve a catalog model or your own weights on GPUs you own, behind an OpenAI-compatible endpoint. ## What you can do - [Chat & reasoning](https://docs.privatemind.com/models.html): Open-weight models with reasoning, long context and vision. - [Files & RAG](https://docs.privatemind.com/files-and-sources.html): Upload PDFs, Word files, spreadsheets and more, and get answers that cite them. - [Tool use & agents](https://docs.privatemind.com/tool-use.html): Let models call your functions and MCP servers, or run agents on a schedule. - [Code execution](https://docs.privatemind.com/chat.html): Sandboxed Python and Node.js environments. The model writes and runs code, then acts on the output. - [API keys & auth](https://docs.privatemind.com/authentication.html): Authenticate, manage keys, and set budgets and rate limits. - [Python SDK](https://docs.privatemind.com/sdk.html): Run training, fine-tuning, image and batch jobs from Python. Installed in every workspace. ## Get started in 3 steps 1. [Create an API key](https://docs.privatemind.com/api-keys.html) on the **API Keys** page in the sidebar. 2. [Pick a model](https://docs.privatemind.com/models.html). The `fast` alias is a good default. 3. Send your first request: **cURL** ```bash curl -s "https://api.privatemind.com/v1/chat/completions" \ -H "Authorization: Bearer $PMIND_KEY" \ -H "Content-Type: application/json" \ -d '{ "model": "fast", "messages": [{"role": "user", "content": "Hello, world."}] }' ``` **Python** ```python from openai import OpenAI client = OpenAI(base_url="https://api.privatemind.com/v1", api_key=PMIND_KEY) resp = client.chat.completions.create( model="fast", messages=[{"role": "user", "content": "Hello, world."}], ) print(resp.choices[0].message.content) ``` **TypeScript** ```typescript import OpenAI from 'openai'; const client = new OpenAI({ baseURL: 'https://api.privatemind.com/v1', apiKey: process.env.PMIND_KEY }); const resp = await client.chat.completions.create({ model: 'fast', messages: [{ role: 'user', content: 'Hello, world.' }], }); console.log(resp.choices[0].message.content); ``` ## Pick your path If you've used the OpenAI API, you already know this one: change the base URL and keep your SDKs and tools. The same key and models also answer the Anthropic [Messages API](https://docs.privatemind.com/messages-api.html), so the Anthropic SDKs and Claude Code work too. So do clients built on the OpenAI [Responses API](https://docs.privatemind.com/responses-api.html), such as Codex CLI and the OpenAI Agents SDK. If you'd rather work in a browser, the PrivateMind app has chat, file uploads, code execution and scheduled automations. ## What's in these docs The sidebar lists every page. Most people start with one of these: - [Quickstart](https://docs.privatemind.com/quickstart.html): your first API call in five minutes - [Models](https://docs.privatemind.com/models.html): the models you can call, live from your environment - [Authentication](https://docs.privatemind.com/authentication.html): bearer token format and key management - [Chat](https://docs.privatemind.com/chat.html): using the web app - [Rate limits & budgets](https://docs.privatemind.com/rate-limits.html): how limits work and what happens when you hit them - [OpenAI SDK](https://docs.privatemind.com/openai-sdk.html): Python, TypeScript, Go, LangChain and other OpenAI clients - [Anthropic SDK](https://docs.privatemind.com/anthropic-sdk.html): the Messages API from the Anthropic Python and TypeScript clients - [Python SDK](https://docs.privatemind.com/sdk.html): training, image generation and batch jobs from Python (`pip install privatemind`)