Unlock Productivity: Create a Full‑Featured AI Coding Workflow on Ubuntu with CC Switch and Ollama

Step 1 – Foundation: Install Ollama and Prepare Models

Ollama serves as the local AI inference engine. Install it with the official one‑liner: curl -fsSL https://ollama.com/install.sh | sh Configure the systemd service so Ollama listens on all interfaces: sudo systemctl edit ollama.service In the editor add:

[Service]
Environment="OLLAMA_HOST=0.0.0.0:11434"
Environment="OLLAMA_ORIGINS=*"

Reload and restart the service:

sudo systemctl daemon-reload
sudo systemctl restart ollama

Download and run a programming‑oriented model. The guide recommends DeepSeek‑Coder‑V2 or Qwen2.5‑Coder (e.g., the 32B variant requiring ≥24 GB VRAM):

# Qwen2.5‑Coder 32B
ollama run qwen2.5-coder:32b
# or DeepSeek‑R1 14B
ollama run deepseek-r1:14b

Step 2 – Tools: Install AI CLI Clients

Install the command‑line front‑ends that normally talk to cloud APIs. Ensure Node.js v18+ and npm are present.

Claude Code (Anthropic)

npm install -g @anthropic-ai/claude-code

Codex CLI (OpenAI compatible)

npm install -g @openai/codex

Gemini CLI (Google)

npm install -g gemini-chat-cli

Running these commands would normally prompt for an API key; the next step intercepts that traffic.

Step 3 – Hub: Install and Configure CC Switch

CC Switch (Claude Code Switch) is an open‑source proxy router that can rewrite requests from the above CLIs to any OpenAI‑compatible backend, such as Ollama.

Install CC Switch

npm install -g @songhe/cc-switch
# or download the Linux binary from GitHub Releases
# wget https://github.com/farion1231/cc-switch/releases/download/v3.10.0/CC-Switch-v3.10.0-linux-amd64.tar.gz

Add an Ollama configuration

# Create a new profile named local-ollama
ccs new local-ollama

During the interactive setup, provide:

Provider Type: OpenAI Compatible Base URL: http://localhost:11434/v1 API Key: ollama (placeholder)

Model: the name of the model you started, e.g.,

qwen2.5-coder:32b

Start the proxy

# Switch to the new profile
ccs switch local-ollama
# Launch proxy mode (required by some clients)
ccs proxy start

CC Switch now runs a background forwarding service and sets environment variables such as https_proxy, HTTP_PROXY, and CLAUDE_BASE_URL so the CLI tools connect to the local Ollama instance.

Final Test – Verify the Workflow

Run the Claude Code CLI: claude The prompt is processed locally; no login to Anthropic is needed. Example prompt: >> 帮我用 Python 写一个贪吃蛇游戏 Ollama performs the inference, and Claude Code returns the generated Python code within seconds.

Advanced Tips

Mixed profiles: Use ccs switch to toggle between a local DeepSeek profile and a cloud‑based Claude‑3.5 profile, keeping sensitive data local while delegating heavy reasoning to the cloud.

Hot model swapping: Change the model field in CC Switch configuration to instantly switch Claude CLI to DeepSeek or Llama backends.

Conclusion

By combining Ubuntu, Ollama, and CC Switch, you can enjoy the polished interfaces of major AI services while retaining full control over data and eliminating costly API subscriptions.