How to Use GLM-5 for Free with Ollama

Description and use

GLM‑5 can be run locally via Ollama, avoiding cloud costs and giving full control over the model and data.

Problem

Run the large language model GLM‑5 locally without cloud costs or data transfer.

Outcome

GLM‑5 runs locally with minimal setup, eliminating cloud costs and providing full control over model behavior and data privacy.

Workflow

• Install Ollama from the official website and verify with ollama --version
• Start the Ollama server with ollama serve if it does not launch automatically
• Pull the GLM‑5 model with ollama pull glm-5:cloud
• Run an interactive session using ollama run glm-5:cloud
• Use the OpenAI‑compatible REST endpoint at http://localhost:11434/v1 for programmatic access, e.g., with curl or the OpenAI SDK
• Optionally create a custom Modelfile to set system prompt, temperature, and context length, then build with ollama create my-glm5 -f Modelfile and run with ollama run my-glm5
• Test and debug API calls visually with Apidog by configuring base URL, endpoint, headers, and request body
• Generate SDK code or automated tests in Apidog for structured workflow validation

Evidence

GLM‑5 from Z.ai is a frontier-level open‑source model you can now run through Ollama — complex reasoning, software engineering, and long‑horizon agentic workflows, entirely on your own hardware.

Source

Download the installer for your platform from the official Ollama website. On macOS or Linux, run the installation command from the site in a terminal; on Windows, execute the downloaded .exe.

Source

Pull the model: ollama pull glm-5:cloud

Source

The official ollama Python library works as well.

Source