Command-line coding assistant
A coding assistant that lives in your terminal.
ZynOps reads your project, edits files, and runs commands from the shell you already work in. Sign in once from the browser, then point it at whichever model provider you want to use.
$ npm install -g @zynopssolutions/zynops-cliFree plan, no card required · v1.1.3
$ zynops loginOpening ZynOps login in your browser...If the browser doesn't open, visit: https://cli.zynopssolutions.online/cli/authorize?code=QK7F-2WD9 Enter code QK7F-2WD9 if prompted. Waiting for authorization... (expires in 10 minutes)Login successful. Connected as you@example.com. $ zynops# starts an interactive session in the current directory- Runtime
- Node.js 18+
- Platforms
- macOS · Linux · Windows
- Models
- Any OpenAI-compatible API
- Plans
- $0 – $20 / month
How it works
Three commands from install to first edit.
- 01
Install the CLI
One global npm package. It ships a single binary named zynops and nothing else.
$ npm install -g @zynopssolutions/zynops-cli - 02
Sign in from the browser
The CLI prints a short device code and opens your browser to approve it. Approve once and the session token is written to your machine — never into the repository.
$ zynops login - 03
Run it in a project
Start an interactive session in whatever directory you are working in, or pass a command straight through for scripted use.
$ cd my-project && zynops
Bring your own model
Not tied to one vendor.
ZynOps talks to any endpoint that speaks the OpenAI chat API. Use a hosted provider, or keep the whole loop on your own hardware with Ollama or LM Studio and send nothing off the machine.
Provider setup in the docs| Provider | Endpoint | Runs |
|---|---|---|
| OpenAI | api.openai.com | Hosted |
| DeepSeek | api.deepseek.com | Hosted |
| Ollama | localhost:11434 | On your machine |
| LM Studio | localhost:1234 | On your machine |
What it does
It works in the repository, not in a chat window.
Reads the codebase first
It opens files, follows imports, and searches the tree for what it needs, so you are not pasting context into a chat box by hand.
Edits files in place
Changes land in your working tree across as many files as the task needs. Review them with git diff like any other change, and revert them the same way.
Runs your commands
It can run the build, the test suite, or a one-off script, read what came back, and correct itself before handing the work over.
Interactive or scripted
Run zynops on its own for a session in the terminal, or pass arguments through and use it from a shell script or a git hook.
Usage without leaving the shell
zynops usage and zynops account print your plan, today’s request count, and when the counter resets. The same numbers appear in the dashboard.
Config stays local
Settings and your session token live in ~/.zynops, or in your operating system keychain where one is available. Nothing is written into the project.
Pricing
Daily limits, priced in single dollars.
| Plan | Per month | Requests / day | Credits / day |
|---|---|---|---|
| Free | Free | 3 | 3 |
| Starter | $2 | 8 | 10 |
| Basic | $5 | 20 | 25 |
| Pro | $10 | 40 | 50 |
| Max | $20 | 100 | 150 |
A request spends 1 credit for a basic task, 3 for a normal one, 5 for an advanced one, and 10 for long-context work — so the credit limit is usually what binds, not the request count. Both counters reset daily at 00:00 UTC.
Limits apply to requests that ZynOps serves. When you point the CLI at your own provider key or at a model running locally, that traffic is billed by your provider instead.
One install, and it is in every project on your machine.
$ npm install -g @zynopssolutions/zynops-cli