<picture>


</picture>

Askimo

Askimo is a powerful, pluggable command-line chat assistant that lets you talk to AI models like OpenAI, Ollama, X AI, and more — right from your terminal. It’s built for convenience with session-based configuration that’s saved locally, and it’s easy to extend with a simple plugin-style system.

Ask anything, switch models, and customize your AI experience — all from your terminal.

Features

• Talk to many LLMs in one place — Use OpenAI, X AI, Ollama, and more without switching tools. The pluggable design makes it easy to add future providers as they appear.

• Switch models instantly – Change providers or models on the fly with a single command.

• Get answers in real time – Responses stream as they’re generated, so you don’t have to wait.

• Stay configured between sessions – Your models, providers, and API keys are saved locally and auto-loaded next time.

• Customize your experience – Tune temperature, top_p, max_tokens, and other parameters per provider.

• Pluggable architecture – Extend Askimo with your own models and commands.

• Work in a familiar REPL – Interactive terminal environment with built-in commands for setup, copying, and more.

• Command piping – Send output from other shell commands into Askimo (git log | askimo) and get AI-generated summaries or transformations.

• Portable and fast – Compiled as a native binary with GraalVM for instant startup.

• Runs anywhere – Fully cross-platform on macOS, Linux, and Windows.

• File system integration – Read and process local files directly from Askimo, turning their contents into AI prompts.

Coming soon

• Custom prompt shortcuts – Define reusable commands like :release_notes <commits_file> to instantly generate release notes from a list of commits.</commits_file>

• Scripting mode – Run Askimo in non-interactive mode inside scripts or CI/CD pipelines.

Demo

• Piping commands & switching providers in Askimo

• Interacting with the local file system in Askimo

💬 Simple Web Chat (Local Usage)

Askimo isn’t only for the terminal — you can also start a lightweight local web chat UI if you prefer a browser interface.
This feature is designed for quick testing or personal use, not for production deployment.

• Start Askimo web server

askimo --web

Then open your browser to the URL printed in the console (look for Web server running at http://127.0.0.1:8080). If port 8080 is busy, Askimo will pick the next free port—use the exact address shown in that log line.
The web UI supports real-time streaming responses and Markdown rendering.

⚠️ Important: Before running askimo web, you must finish setting up your AI provider (e.g., Ollama, OpenAI) and select a model using the Askimo CLI.
The web version is currently a simple chat page — it does not support configuring providers, models, or AI parameters. All configuration must be done beforehand via the CLI.

Installation

You can install Askimo in three ways today:

1. Homebrew (macOS/Linux)

If you’re on macOS or Linux, the easiest way is via Homebrew

brew tap haiphucnguyen/askimo
brew install askimo

Once installed, you can run:

askimo

to start the CLI.

To update Askimo later:

brew upgrade askimo

2. Download Release Binaries

Prebuilt binaries for macOS, Linux, and Windows are available on the Releases page.

1. Download the archive for your operating system.
2. Extract it.
3. Move the askimo (or askimo.exe on Windows) binary to a directory on your $PATH.

Example (Linux/macOS):

mv askimo /usr/local/bin/

3. Build from Source

If you prefer to build locally:

git clone git@github.com:haiphucnguyen/askimo.git
cd askimo
./gradlew nativeCompile

The binary will be located at build/native/nativeCompile/askimo

Run it directly:

./build/native/nativeCompile/askimo

Coming Soon

We plan to support installation via:

• Scoop (Windows)

• Linux package managers (APT, DNF)

Stay tuned for updates!

Available Commands

➡ View the full command reference »
Includes detailed usage, options, and examples for each command.

💡 Note: Some providers (such as OpenAI, X AI, etc.) require an API key.
Make sure you create and configure the appropriate key from your provider’s account dashboard before using them.

Extending Askimo

Askimo is designed to be pluggable, so you can tailor it to your needs:

• Add a new chat model – Integrate any AI provider by following the guide in docs/creating-new-chat-model.md

• Create a new command – Add custom CLI commands to automate tasks or build integrations. See docs/creating-new-command.md.

Contributing

• Fork & clone the repo

• Create a feature branch

• Open a PR