modelstudioai/cli

Official Model Studio CLI（阿里云百炼 CLI）built for AI Agent frameworks, exposing models, search, multimodal, and workflow capabilities as structured tool calls.

Aliyun Model Studio CLI (bailian‑cli)

What it is – A command‑line interface that lets you interact with Alibaba Cloud’s Model Studio (aka DashScope) AI platform. It wraps the platform’s REST APIs into easy‑to‑type commands, so you can generate text, images, video and speech, run inference, fine‑tune models, manage datasets, and control cloud‑hosted “Managed Agents” without writing any HTTP code.

Who it’s for – Developers, data scientists, and AI‑agent builders who already have an Alibaba Cloud account and want a terminal‑based way to call the platform’s multimodal models and management features.

Core capabilities

Area	What you can do from the CLI
Generation	Text, image, video, and speech creation; editing and reference‑based generation (e.g., image‑to‑video).
Understanding	Ask questions about images, PDFs, audio, or long videos – the CLI sends the asset to the model and returns answers.
App orchestration	Deploy, preview, and run Managed Agents and workflows that combine knowledge bases, memory, web search, and other tools.
Training & deployment	Validate datasets, start fine‑tuning jobs, and publish a model as an endpoint.
Account ops	Log in, view usage/quota, manage team seats, increase rate limits, browse the model marketplace.
Plan onboarding	Attach a Token‑Plan subscription to the CLI so agents can consume paid tokens automatically.

Note – The orchestration, training, and account‑management commands currently work only for accounts on the China‑site (aliyun.com).

Quick start workflow

Install – either via npm (npm i -g bailian-cli), a one‑line curl/PowerShell installer, or by asking an AI Agent to run the install script for you.
Authenticate – obtain an API key from the DashScope console and run bl auth login --api-key sk‑… (or use OAuth console login, or Alibaba Cloud AK/SK for seat‑management).
Run a task – simply describe what you want to an AI Agent (e.g., “Generate a cat in a spacesuit on Mars and turn it into a video”). The Agent translates the natural‑language request into the appropriate bl commands and executes them.
Manage – use bl config, bl update, and bl auth sub‑commands to switch profiles, change language, or upgrade the tool.

Installation options

Platform	Command
Node (npm)	`npm install -g bailian-cli && bl skill init`
macOS / Linux (script)	`curl -fsSL https://bailian.aliyun.com/cli/install.sh \| bash`
Windows (PowerShell)	`irm https://bailian.aliyun.com/cli/install.ps1 \| iex`
Agent‑assisted	Send the text “Please read https://bailian.aliyun.com/cli/install.md and install the Aliyun Model Studio CLI for me” to your AI Agent.

Example commands

# Generate an image and turn it into a short video
bl gen image "a cat in a spacesuit on Mars" --output cat.png
bl gen video --input cat.png --duration 5s

# Transcribe audio with hot‑word correction
bl speech transcribe audio.wav --hot-words "Mars,SpaceX"

# List your model usage and quota
bl usage show

# Deploy a Managed Agent defined in a JSON config
bl agent deploy --config director.json

Where to learn more

Official site – https://bailian.console.aliyun.com/cli
API docs – https://help.aliyun.com/zh/model-studio/
Chinese README – https://github.com/modelstudioai/cli/blob/main/README.zh.md
Change log – https://github.com/modelstudioai/cli/blob/main/CHANGELOG.md
Community – DingTalk user‑group QR code shown in the repo’s README.

TL;DR

The Aliyun Model Studio CLI (bailian-cli) is a full‑featured terminal client for Alibaba Cloud’s multimodal AI platform. It lets you generate and understand media, fine‑tune and deploy models, and orchestrate cloud‑hosted AI agents—all from simple bl … commands, with optional natural‑language “agent” front‑ends that translate your spoken request into the right CLI calls.