Integration Guide — CC Switch or Manual Setup in 5 Minutes

⚡ CC Switch — for Claude Code / Codex CLI / Gemini CLI users

Best if you switch between official and relay providers, or run multiple coding agents

Download CC Switch

Grab the latest release for your platform from the CC Switch GitHub repository. Small desktop app, zero dependencies.

Add the Waterseller provider

Open CC Switch → Providers → + Add. Fill in: Name: Waterseller · Base URL: https://waterseller.cc · API Key: your sk-... key from the Tokens page.

Pick your models

CC Switch auto-fetches the model catalog — glm-5.3-flash, deepseek-v4-pro, gpt-6-sol, grok-4.7 and 38 more. Set your default.

Switch anywhere, anytime

One click rewrites the config of Claude Code / Codex CLI / Gemini CLI on your machine. Toggle back to official billing whenever you want.

💡 After adding the provider, CC Switch handles config rewrites for you. Toggle once — every CLI on your machine now runs on Waterseller pricing. Toggle back to switch to official billing at any time.

What CC Switch writes (for the curious)

Example: the generated snippet when Waterseller is active:

base_url = "https://waterseller.cc/v1"
api_key  = "sk-...your-waterseller-key..."
model    = "glm-5.3-flash"   # or any of 42 models

🔧 Manual setup — for any OpenAI-compatible tool

Best if you use raw SDKs, LangChain, custom agents, or anything with a base_url field

Get your API key

Point base_url at Waterseller

Set the base URL to https://waterseller.cc/v1 and use your key as the bearer token. That's the only change.

Pick a model

Use any model id from the pricing page — e.g. glm-5.3-flash, deepseek-v4-pro, minimax-m3.1-flash.

Ship it

Streaming, tool calling, JSON mode and vision all work as-is. Monitor usage in your dashboard in real time.

Python (OpenAI SDK)

from openai import OpenAI

client = OpenAI(
    api_key="sk-...",                       # your Waterseller key
    base_url="https://waterseller.cc/v1"    # one-line switch
)

resp = client.chat.completions.create(
    model="glm-5.3-flash",
    messages=[{"role": "user", "content": "Hello!"}]
)
print(resp.choices[0].message.content)

curl

curl https://waterseller.cc/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer sk-..." \
  -d '{
    "model": "glm-5.3-flash",
    "messages": [{"role": "user", "content": "Hello!"}]
  }

Environment variables (any tool)

# Any tool that reads OPENAI_* env vars
export OPENAI_API_KEY="sk-..."
export OPENAI_BASE_URL="https://waterseller.cc/v1"

FAQ

Integration questions.

Which path should I choose?

Using Claude Code, Codex CLI or Gemini CLI and want instant switching between providers? Take CC Switch. Writing your own code or using any tool with a configurable OpenAI base URL? Manual takes 3 lines. Both connect to the same models at the same prices.

Do both methods use the same API key?

Yes — one Waterseller key works everywhere: through CC Switch-managed CLIs, raw SDK calls, curl, LangChain, anything. Create as many keys as you like on the Tokens page.

Is the OpenAI-compatible endpoint fully compatible?

Yes: /v1/chat/completions, /v1/models, streaming SSE, tool calling, JSON mode, vision inputs. If a tool speaks the OpenAI protocol, it works unchanged.

Can I switch between models mid-project?

Change the model id string — that's it. All 42 models share the same key, the same billing pool and the same endpoint.

Connect in two ways:
CC Switch or manual setup