AI Coding Agents: A Practical Guide to Claude Code with Ollama

Ismail Ajjal — Tue, 20 Oct 2026 02:28:33 +0000

Want to try an AI coding agent without sending your project to a hosted model? This guide connects Claude Code to a local Ollama model. Commands are provided for Windows CMD, PowerShell, Linux and macOS.

Our test setup: Windows, 8 GB RAM, Ollama, Claude Code and qwen2.5:1.5b. Results and limitations from this setup are called out separately.

What you’ll do: install the tools → download a model → connect Claude Code → test a real file operation.

1. Install Ollama

Ollama runs the model locally. Use the command for your operating system, or use the official download page for a manual installer.

Windows — PowerShell:

irm https://ollama.com/install.ps1 | iex

Manual installation:
Download and run the Windows installer from ollama.com/download, then reopen CMD.

macOS or Linux — Terminal:

curl -fsSL https://ollama.com/install.sh | sh

Manual installers and Linux-specific instructions: Ollama downloads · Linux manual install.

2. Install Claude Code

Windows — PowerShell:

irm https://claude.ai/install.ps1 | iex

macOS / Linux — Terminal:

curl -fsSL https://claude.ai/install.sh | bash

Official instructions: Claude Code setup.

3. Verify the commands

Close and reopen the terminal, then run:

ollama --version
claude --version

4. Download and test the local model

For this experiment, we used a small coding model to fit a computer with 8 GB RAM:

ollama pull qwen2.5:1.5b

Run the model and ask a simple question to confirm that it responds, then type /bye to exit the model chat.

ollama run qwen2.5:1.5b

Check installed models with:

ollama list

The first download needs an internet connection. After the model is downloaded, inference runs on the local machine. Speed and quality depend on your hardware and the model.

5. Connect Claude Code to Ollama

There are two methods. Start with Method A; use Method B only if you need to configure the connection manually. Do not mix the two configurations unnecessarily.

Method A — Ollama launcher (recommended)

ollama launch claude --model qwen2.5:1.5b

This uses Ollama’s Claude Code integration and selects the local model. If launch is not recognized, update Ollama using the current installer from ollama.com/download, then reopen the terminal.

Method B — Environment variables (manual setup)

Claude Code can be pointed at Ollama’s local API with environment variables. These variables are not required when Method A works. They are a way to configure the manual connection.

Windows CMD — current window only:

set ANTHROPIC_BASE_URL=http://localhost:11434
set ANTHROPIC_AUTH_TOKEN=ollama
set ANTHROPIC_API_KEY=
claude --model qwen2.5:1.5b

Windows PowerShell — current window only:

$env:ANTHROPIC_BASE_URL = "http://localhost:11434"
$env:ANTHROPIC_AUTH_TOKEN = "ollama"
$env:ANTHROPIC_API_KEY = ""
claude --model qwen2.5:1.5b

Linux / macOS — current shell only:

export ANTHROPIC_BASE_URL=http://localhost:11434
export ANTHROPIC_AUTH_TOKEN=ollama
export ANTHROPIC_API_KEY=""
claude --model qwen2.5:1.5b

These commands apply to the terminal session in which you run them. Close that session and the temporary variables disappear.

Do I need to set them every time?

With ollama launch claude: usually no manual variable setup is needed.
With set, $env: or export: yes, set them again in each new terminal session.
Persistent variables are possible, but are not needed for this tutorial. Avoid setx unless you deliberately want Windows variables to persist; it affects future sessions, not the current CMD window.

6. Test a real file operation: Local vs Cloud

After connecting Claude Code to Ollama, we tested two configurations using the same prompt: one with a local model and one with a cloud model.

First, create an empty test folder and open it in the terminal. Then launch the agent with the configuration you want to test.

Test prompt:

Create hello.py containing print("Hello from local Ollama"). Show me the file contents and explain how to run it. Do not modify any other files.

Test A — Local model: `qwen2.5:1.5b`

We ran the task using the local model on our Windows computer with 8 GB RAM.

The model explained how to create the file manually, but it did not create the requested file in our observed test.

Test B — Cloud model: `gemma:31b-cloud`

We then repeated the same task using gemma:31b-cloud.

In this separate test, the agent carried out the requested actions in the workspace, and we could verify the resulting file and its contents.

After each test, verify the file directly in the workspace instead of relying only on the agent's final message.

7. Comparison and limitations

The two tests produced different results. The table below summarizes the main differences between the configurations and what we observed.

Feature	Local — `qwen2.5:1.5b`	Cloud — `gemma:31b-cloud`
Where it runs	On our computer through Ollama	On remote infrastructure
Local resources	Uses local RAM and CPU/GPU	Most inference resources are remote
Internet	Needed for the initial download; usually not for local inference	Generally required
Privacy	Prompts can stay local, depending on the tools and configuration	Requests are sent to the provider
Result in our file test	Explained how to create the file, but did not create it	Carried out the requested actions in our test
Main limitation	Performance and reliability depend on the model and hardware	Depends on the service, model, network and task

What did we learn?

On our Windows computer with 8 GB RAM, the local model was usable for basic chat but did not consistently behave as a reliable coding agent. Some tool calls were malformed, the workflow was slow, and heavier local models also showed significant slowdowns in our tests.

The cloud model completed the file operation in our separate test. This suggests that a cloud model can be a practical alternative for demanding tasks on a low-resource computer.

However, these results come from our particular tests. They are not a controlled benchmark and do not prove that every cloud model will outperform every local model.

Conclusion

Ollama makes it possible to experiment with Claude Code using a local model, without relying on a hosted model API for local inference.

Our comparison highlights an important lesson: running a model locally does not automatically make an AI coding agent fast or reliable.

Start with a small task, verify every file operation, and choose a model that fits your hardware and requirements. Consider privacy, Internet access and performance before deciding between a local and a cloud configuration.

DEV Community: Ismail Ajjal

AI Coding Agents: A Practical Guide to Claude Code with Ollama

1. Install Ollama

2. Install Claude Code

3. Verify the commands

4. Download and test the local model

5. Connect Claude Code to Ollama

Method A — Ollama launcher (recommended)

Method B — Environment variables (manual setup)

6. Test a real file operation: Local vs Cloud

Test A — Local model: `qwen2.5:1.5b`

Test B — Cloud model: `gemma:31b-cloud`

7. Comparison and limitations

What did we learn?

Conclusion

References

DEV Community: Ismail Ajjal

AI Coding Agents: A Practical Guide to Claude Code with Ollama

1. Install Ollama

2. Install Claude Code

3. Verify the commands

4. Download and test the local model

5. Connect Claude Code to Ollama

Method A — Ollama launcher (recommended)

Method B — Environment variables (manual setup)

6. Test a real file operation: Local vs Cloud

Test A — Local model: qwen2.5:1.5b

Test B — Cloud model: gemma:31b-cloud

7. Comparison and limitations

What did we learn?

Conclusion

References

Test A — Local model: `qwen2.5:1.5b`

Test B — Cloud model: `gemma:31b-cloud`