Apis AI – Wiki

Welcome to the documentation for Apis AI, the native desktop client for local and cloud LLMs. Here you'll find step-by-step guides for installation, providers, MCP servers, local tools, scripts and automated workflows.

Topics at a Glance

Installation

APT repository, .deb, tarball, Windows bundle and JAR – including prerequisites.

Getting Started

Understand the interface, start your first chat and get to know the most important controls.

Providers & Models

Ollama, Claude, Gemini, OpenRouter, DeepSeek, Puter and your own OpenAI-compatible endpoints.

Chat Features

Artifacts, Markdown, charts, file references, tool calls, Git-style file diffs, questions from the AI, compression and rewind.

Speech Input

Dictation and hands-free voice chat with local whisper.cpp or an OpenAI-compatible service.

Understanding & Using MCPs

No technical knowledge required: what MCPs are, what they do for you and how to use them in chat.

Setting Up MCP Servers

Reference: local stdio and remote servers, library, OAuth and troubleshooting.

Tools & Security

Shared folders, terminal sandbox with whitelist/blacklist and delete confirmation.

Scripts

Share your own local scripts, run them with arguments and make them available to the AI as a tool.

Rules & Skills

Persistent instructions for all chats, project rules from the AGENTS.md file and reusable capabilities for when you need them.

Workflows

Visual AI flows made of agent, script, sub-workflow and metadata nodes.

Background & Schedules

System tray, autostart and cron-based workflow schedules.

Settings

All tabs: providers, system & prompts, extensions, parameters and appearance.

FAQ & Troubleshooting

Common questions about models, context window, sandbox and storage locations.

Quick Start

Install – set up the package or repository (see Installation).
Create a provider – add an endpoint and API key under Settings, then Providers & Models (see Providers & Models).
Chat – pick a model in the chat tab and get started (see Getting Started).
Extend – add MCPs and use them in chat (see Understanding & Using MCPs).
Automate – build workflows and run them on a schedule (see Workflows).

About This Wiki This documentation describes Apis AI from version 0.1.x and is maintained alongside the app. Questions or feedback? Write to us at support@apisclient.com.