Generic Client Setup

Any OpenAI-compatible client works out of the box. Below are the exact settings for popular tools.

Common settings (same for every client)

API URLhttps://token.bitbeam.cn/v1
API Keysk-tr-••••••Create one under API Keys

Use the model ID from the Model Market. Some clients require adding models manually instead of fetching the list.

Cherry Studio

OpenAI

Cross-platform desktop client with multi-provider and knowledge base support.

SettingValue
Provider typeOpenAI
API Hosthttps://token.bitbeam.cn/v1
API Keysk-tr-xxxxxx

Add a provider of type "OpenAI", then click Manage to fetch the model list automatically.

ChatBox

OpenAI

Lightweight desktop / mobile client.

SettingValue
Provider typeOpenAI API Compatible
API Host / Endpointhttps://token.bitbeam.cn/v1
API Keysk-tr-xxxxxx

LobeChat

OpenAI

Modern open-source AI chat framework, self-hostable.

SettingValue
OPENAI_PROXY_URLhttps://token.bitbeam.cn/v1
OPENAI_API_KEYsk-tr-xxxxxx

Set as environment variables when self-hosting, or fill them under Settings → Language Model on the hosted version.

NextChat

OpenAI

One-click deployable lightweight web client.

SettingValue
BASE_URLhttps://token.bitbeam.cn
OPENAI_API_KEYsk-tr-xxxxxx

NextChat appends /v1 itself, so BASE_URL must be the domain only — do not include /v1.

Open WebUI

OpenAI

Self-hosted web UI, often used alongside local models.

SettingValue
OPENAI_API_BASE_URLhttps://token.bitbeam.cn/v1
OPENAI_API_KEYsk-tr-xxxxxx

Dify

OpenAI

LLM app orchestration platform — bring these models into workflows and agents.

SettingValue
Provider typeOpenAI-API-compatible
API endpoint URLhttps://token.bitbeam.cn/v1
API Keysk-tr-xxxxxx
Model nameA model ID from the Model Market, e.g. glm-5.2

Add under Settings → Model Provider → OpenAI-API-compatible. Each model must be added individually.

Don't see your tool?

Any tool that lets you set a custom OpenAI endpoint will work with the common settings above. Claude Code, Codex and Cline have dedicated step-by-step guides in the sidebar.