API Profiles

Configure custom AI providers with your own API keys or local models. Full control over model selection, deployment, and cost optimization.

Use ccs glm for supported Z.AI API profiles and ccs km for reasoning-first Kimi API profiles. Legacy ccs glmt remains compatibility-only and is not recommended for new setup.

Available Providers

Provider	Type	Model	Context	Best For
Anthropic Direct	API Key	Claude 4.5	200K	Official Claude access with direct API key
OpenRouter	API Key	300+ models	Varies	Unified access to multiple providers
Novita AI	API Key	`deepseek/deepseek-v3.2`	Varies	Anthropic-compatible hosted endpoint with CCS preset
llama.cpp	Local	GGUF models	Varies	Private, offline local inference
Ollama	Local	Open-source	Varies	Local or cloud-hosted open models
Alibaba Coding Plan	API Key	Qwen3 Coder	1M tokens	Cost-effective coding, Alibaba Cloud
Z.AI GLM (`glm`)	API Key	GLM 5	Standard	Cost-optimized bulk tasks
Kimi for Coding (`km`)	API Key	`kimi-k2-thinking-turbo`	Varies	Moonshot direct API key profile
Custom	API Key	Any	Varies	OpenAI-compatible endpoints

Anthropic Direct API

Available since v7.52.0

Use your own Anthropic API key for direct access to Claude models without going through CLIProxy. Perfect for users with existing Anthropic subscriptions.

Quick Setup

ccs api create --preset anthropic
# Enter your API key (sk-ant-...) when prompted

Or via dashboard: API Profiles > Create Profile > Anthropic Direct

Full Anthropic Direct Guide

Learn about API key setup, available Claude models, and cost tracking

llama.cpp (Local GGUF Inference)

Available since v7.52.0

Run GGUF-quantized models locally via llama.cpp server with zero API costs and complete privacy.

Quick Setup

# 1. Start llama.cpp server (in separate terminal)
./server -m /path/to/model.gguf --host 0.0.0.0 --port 8080

# 2. Create CCS profile
ccs api create --preset llamacpp

# 3. Use it
ccs llamacpp "your prompt"

Full Llama.cpp Guide

Learn about model selection, performance tuning, and GPU acceleration

OpenRouter (300+ Models)

OpenRouter provides access to 300+ AI models through a unified API. CCS includes first-class OpenRouter support with interactive model picker and tier mapping.

Quick Setup

ccs api create --preset openrouter

Or via dashboard: API Profiles > Create Profile > OpenRouter

Full OpenRouter Guide

Learn about model selection, tier mapping, and cost optimization

Novita AI

Available since v7.54.0

Novita AI is available as a built-in API preset for Anthropic-compatible routing. CCS ships the correct base URL and a default profile name so you can get from API key to prompt quickly.

Quick Setup

ccs api create --preset novita
ccs novita "triage this bug report"

Or via dashboard: API Profiles > Create Profile > Novita AI

Full Novita Guide

Learn about preset setup, manual configuration, and model overrides

Alibaba Coding Plan

Available since v7.51.0

Alibaba Cloud Coding Plan provides Qwen3 Coder models through an Anthropic-compatible endpoint.

Quick Setup

ccs api create --preset alibaba-coding-plan
# or use aliases:
ccs api create --preset alibaba
ccs api create --preset acp

Full Alibaba Coding Plan Guide

Learn about Qwen3 Coder models, pricing, and Coding Plan key setup

Profile Lifecycle Management

Profile lifecycle features available since v7.52.0

CCS provides CLI commands for discovering, copying, exporting, and importing API profiles.

Discover Orphaned Profiles

Find API profile settings files that aren’t registered in config:

ccs api discover
# Lists orphaned .settings.json files not in config

ccs api discover --register
# Registers discovered orphan profiles

Copy Profiles

Duplicate an existing profile with a new name:

ccs api copy glm glm-test
# Copies glm settings to glm-test (same settings, new name)

Export Profiles

Export a profile to a portable JSON file (with or without API keys):

# Export with API key (secure storage only)
ccs api export glm --out glm-backup.json --include-secrets

# Export without API key (safe to share/backup)
ccs api export glm --out glm-backup.json

Import Profiles

Import profiles from exported JSON files:

# Import from backup
ccs api import glm-backup.json

# Import into specific profile name
ccs api import glm-backup.json --name glm-restored

Quick Start

1. Open Dashboard

ccs config

2. Add API Profile

Navigate to “API Profiles” tab and add your provider:

3. Use Profile

ccs glm "your prompt"
ccs km "analyze this large codebase"
ccs anthropic "explain this code"  # Direct API
ccs novita "debug this stack trace"
ccs llamacpp "local analysis"      # Local llama.cpp

Manual Configuration

GLM Profile

Create ~/.ccs/glm.settings.json:

{
  "env": {
    "ANTHROPIC_BASE_URL": "https://api.z.ai/api/anthropic",
    "ANTHROPIC_AUTH_TOKEN": "your_glm_api_key",
    "ANTHROPIC_MODEL": "glm-5"
  }
}

Kimi for Coding Profile

Create ~/.ccs/km.settings.json:

{
  "env": {
    "ANTHROPIC_BASE_URL": "https://api.kimi.com/coding/",
    "ANTHROPIC_AUTH_TOKEN": "your_kimi_api_key",
    "ANTHROPIC_MODEL": "kimi-k2-thinking-turbo"
  }
}

Register in Config

Add to ~/.ccs/config.json:

{
  "profiles": {
    "glm": "~/.ccs/glm.settings.json",
    "km": "~/.ccs/km.settings.json",
    "default": "~/.claude/settings.json"
  }
}

Custom OpenAI-Compatible Endpoints

Any OpenAI-compatible API can be configured:

{
  "env": {
    "ANTHROPIC_BASE_URL": "https://your-endpoint.com/v1",
    "ANTHROPIC_AUTH_TOKEN": "your-api-key",
    "ANTHROPIC_MODEL": "your-model-name"
  }
}

Use Cases

Budget Conscious

Use GLM for routine tasks at ~$0.50/1M tokens

Long Documents

Kimi for Coding via ccs km

Cost Optimization

Mix Claude (quality) + GLM (volume)

Enterprise

Custom endpoints with your own API keys

Settings Format

All env values MUST be strings (not booleans/objects) to prevent PowerShell crashes.

{
  "env": {
    "ANTHROPIC_BASE_URL": "https://...", // String
    "ANTHROPIC_AUTH_TOKEN": "your-key", // String
    "ANTHROPIC_MODEL": "model-name" // String
  }
}

Hybrid Workflow

Most power users combine subscription + API:

# Terminal 1: Strategic planning with Claude
ccs "Design the authentication architecture"

# Terminal 2: Bulk implementation with GLM
ccs glm "Implement user service based on the design"

# Terminal 3: Code review with Gemini (free)
ccs gemini "Review the implementation for security issues"

Getting Started

Providers

Features

Tutorials

​API Profiles

​Available Providers

​Anthropic Direct API

​Quick Setup

Full Anthropic Direct Guide

​llama.cpp (Local GGUF Inference)

​Quick Setup

Full Llama.cpp Guide

​OpenRouter (300+ Models)

​Quick Setup

Full OpenRouter Guide

​Novita AI

​Quick Setup

Full Novita Guide

​Alibaba Coding Plan

​Quick Setup

Full Alibaba Coding Plan Guide

​Profile Lifecycle Management

​Discover Orphaned Profiles

​Copy Profiles

​Export Profiles

​Import Profiles

​Quick Start

​1. Open Dashboard

​2. Add API Profile

​3. Use Profile

​Manual Configuration

​GLM Profile

​Kimi for Coding Profile

​Register in Config

​Custom OpenAI-Compatible Endpoints

​Use Cases

Budget Conscious

Long Documents

Cost Optimization

Enterprise

​Settings Format

​Hybrid Workflow

API Profiles

Available Providers

Anthropic Direct API

Quick Setup

llama.cpp (Local GGUF Inference)

Quick Setup

OpenRouter (300+ Models)

Quick Setup

Novita AI

Quick Setup

Alibaba Coding Plan

Quick Setup

Profile Lifecycle Management

Discover Orphaned Profiles

Copy Profiles

Export Profiles

Import Profiles

Quick Start

1. Open Dashboard

2. Add API Profile

3. Use Profile

Manual Configuration

GLM Profile

Kimi for Coding Profile

Register in Config

Custom OpenAI-Compatible Endpoints

Use Cases

Settings Format

Hybrid Workflow