Question 1

What is Dragon Code?

Accepted Answer

Dragon Code is a terminal-native AI coding agent (the `dragon` command) wired to Dragon inference at api.mws.run. It runs trillion-parameter models with a 10M-token addressable context and indexes your whole repo, pulling the most relevant code in on every turn. Every claim is grounded in real file:line references.

Question 2

How do I install it, and which platforms are supported?

Accepted Answer

One command for your platform. macOS or Linux: curl -fsSL https://raw.githubusercontent.com/VELLORAAI/dragoncode-public-dist/main/install | bash. Windows (PowerShell): irm https://raw.githubusercontent.com/VELLORAAI/dragoncode-public-dist/main/install.ps1 | iex. No WSL or Ubuntu setup is required on Windows; PowerShell alone installs it. Then open a new terminal, run `dragon --version` to confirm, and run `dragon` in any project to paste your API key. Prerequisites: none on macOS or Windows; Linux just needs curl and tar (most distros already have them). It runs natively on macOS (Apple Silicon and Intel), Linux (x64 / arm64 glibc; Alpine/musl not yet), and Windows 10/11 (x64; ARM64 runs the x64 build under emulation). Git Bash and WSL also work.

Question 3

How does the 10M-token context work?

Accepted Answer

Two things working together. The model brings a very large context window (sparse attention). Dragon Code indexes your whole repo and pulls the most relevant code into that window every turn. You get up to 10M tokens addressable, with the right code inside the window. Nothing important is truncated or forgotten as your repo grows.

Question 4

Do you upload or train on my code?

Accepted Answer

To make whole-codebase retrieval work, Dragon Code indexes your working repo and uploads it to the gateway. That index is owner-scoped: only your API key can read it. Secret files like .env, keys, and credentials are filtered out automatically, before upload and again on our end. You can turn indexing off entirely, delete your index any time, and we never train on your code.

Question 5

What are the different Dragon lanes?

Accepted Answer

Four lanes, selected inline or via the picker: Dragon (default, fast and balanced for everyday coding), Tiny (cheapest, append /tiny), Large (a bigger everyday lane, append /large), and Max (the highest-capability lane for the hardest tasks, append /max). General work stays on Dragon.

Question 6

What is Advisor?

Accepted Answer

Advisor runs automatically, no setup needed. Your session stays on the fast Dragon model, and when it hits a call worth getting right, it hands that one question to a flagship frontier model, takes the answer, and keeps building. You pay frontier rates for that single call, not your whole session.

Question 7

How much does it cost?

Accepted Answer

Plans start at $9/month for Noob (10M monthly tokens) and scale through Starter ($29 / 35M), Builder ($79 / 100M), Pro ($249 / 250M), and Supermax ($750 / 1B). Every plan is a monthly Dragon token pool with no rolling-window lockouts and no surprise bills. If you run past your pool, you can top up with pay-as-you-go credits.

Question 8

Can I use Dragon inference from my own tools?

Accepted Answer

Yes. The Dragon API is a drop-in replacement for the Anthropic and OpenAI APIs. Set your baseURL to https://api.mws.run/v1, add an MWS API key, and your existing SDK, streaming, and tool calls work unchanged. API usage is pay-as-you-go on credits. Get a key at mws.run.

Checking system status…

Components

Past Incidents