Wally

Run open models on your own machine, or hosted when the job outgrows it.

Chat, vision, speech and embeddings from one terminal command. Local models never leave your device. Hosted ones run on RunAnywhere Cloud and are billed against your own credit.

Install

macOS (Apple Silicon) and Linux (x86-64):

curl -fsSL https://raw.githubusercontent.com/RunanywhereAI/wally/main/install.sh | sh

Windows:

irm https://raw.githubusercontent.com/RunanywhereAI/wally/main/install.ps1 | iex

Run a model on your machine

You don't need an account or a key, and nothing leaves the machine.

wally pull qwen3            # download
wally run qwen3             # chat
wally run qwen3 "Hello"     # one answer and exit
wally serve qwen3           # OpenAI-compatible API on :8080 (macOS, Linux)

wally models list --all shows everything you can pull. Any Hugging Face GGUF works too:

wally pull hf.co/Qwen/Qwen3-0.6B-GGUF/Qwen3-0.6B-Q8_0.gguf

Use a hosted model in your coding agent

Sign in once. The terminal shows a code and you approve it in the browser, so you never paste a key.

wally login
wally opencode --cloud -m glm-5.3-flash
wally claude-code -m glm-5.3-flash
wally usage                 # credit left and recent spend

Hosted models today are glm-5.3-flash, qwen3.8-27b and gemma-4. The same commands work with a model on your machine, and claude-desktop, clion and rustrover are wired the same way.

Commands you'll use

wally run	chat, or one answer with a prompt
wally pull / wally rm	download or delete a model
wally list	models on this machine
wally serve	OpenAI-compatible API
wally login / wally usage	sign in, check credit
wally opencode / wally claude-code	start a coding agent on a model

wally --help and wally <command> --help cover the rest, including vision, speech, embeddings and image generation.

Build from source

Needs a built C++ desktop kit, not the SDK source. CONTRIBUTING.md has the steps.

More

• Engines and platforms: what runs where and how wally picks
• Models: the full catalog
• Editors and hosted models: how each tool is wired, where your session lives
• docs.runanywhere.ai · Discord · Hugging Face

MIT. See LICENSE.