Aliyun Bailian(百炼) for OpenClaw#

Call Alibaba Cloud Bailian (百炼) LLM models via the DashScope API.

Setup#

Get your API key from: https://bailian.console.aliyun.com/

Set the environment variable:

bash

export ALIYUN_BAILIAN_API_KEY="your-api-key"
export ALIYUN_BAILIAN_API_HOST="https://coding.dashscope.aliyuncs.com/apps/anthropic"  # optional, default provided

CLI Commands#

bash

python -m scripts chat --model qwen3.5-plus --messages '[{"role": "user", "content": "Hello"}]'
python -m scripts translate --text "Hello" --target-lang zh
python -m scripts models

Commands Overview#

CommandWhat it doeschatGeneral-purpose chat completion with Qwen, GLM, Kimi, MiniMax modelstranslateTranslate text between languagesmodelsList all available models

Chat#

General-purpose chat completion via DashScope Anthropic API.

Usage#

bash

python -m scripts chat --model qwen3.5-plus --messages '[{"role": "user", "content": "Hello"}]'

Parameters#

• --model (string, optional) — Model name. Default: qwen3.5-plus
• --messages (string, required) — JSON array of {role, content}. Roles: system, user, assistant
• --temperature (float, optional) — Sampling temperature 0-1. Default: 0.7
• --max-tokens (integer, optional) — Max tokens to generate. Default: 2048
• --stream (boolean, optional) — Enable streaming. Default: false

Available Models#

Flagship: qwen3.5-plus, qwen3-max-2026-01-23

Coder: qwen3-coder-next, qwen3-coder-plus

Other: glm-5, glm-4.7, kimi-k2.5, MiniMax-M2.5

Translate#

Translate text between languages using the LLM.

Usage#

bash

python -m scripts translate --text "Hello" --target-lang zh
python -m scripts translate --text "你好" --target-lang en --source-lang zh

Parameters#

• --text (string, required) — Text to translate
• --target-lang (string, optional) — Target language code. Default: en
• --source-lang (string, optional) — Source language code. Default: auto

Supported Languages#

en (English), zh (Chinese), ja (Japanese), ko (Korean), es (Spanish), fr (French), de (German), ru (Russian), ar (Arabic), pt (Portuguese), it (Italian), th (Thai), vi (Vietnamese), id (Indonesian)

Python Usage#

python

from scripts import chat, translate

# Chat
result = chat(messages=[{"role": "user", "content": "Hello"}], model="qwen3.5-plus")

# Translate
result = translate(text="Hello", target_lang="zh")

Response Format#

json

{ "success": true, "result": {...} }
{ "success": false, "error": "error message" }

Notes#

• Default API host: https://coding.dashscope.aliyuncs.com/apps/anthropic (set via ALIYUN_BAILIAN_API_HOST)
• The API uses Anthropic-compatible format — messages are converted to Anthropic API format internally
• Coding Plan API key ( sk-sp-xxx) is different from regular DashScope API key
• For reasoning tasks: try qwen3-max-2026-01-23 or kimi-k2.5
• For coding tasks: use qwen3-coder-next or qwen3-coder-plus

Learn More#

• Full API documentation: references/API.md
• Available models: assets/models.json