Ollama configuration

Ollama allows developers to generate commit messages using a Large Language Model (LLM) locally. This helps keep your codebase private by preventing code changes from being sent to external third-party services.

In order to use Ollama, you need to have it installed and running on your machine. For more information, see the Ollama website or the Ollama documentation.

Available models

The following models have been tested and are supported. For more information about all available models, see the Ollama documentation.

Custom models

We're working on a compact, fine-tuned model that will bring fast, accurate commit messages to everyone.

Ollama models

Name
llama3.2:3b
Description
Meta's Llama 3.2, 3B parameters
Name
mistral:7b
Description
Mistral AI's Mistral, 7B parameters
Name
gemma:7b
Description
Google DeepMind's Gemma, 7B parameters
Name
qwen2.5:7b
Description
Alibaba's Qwen 2.5, 7B parameters

Compatible providers

You can use different providers that provide Ollama compatible API by changing the endpoint in the config file.

providers:
  ollama:
    endpoint: http://localhost:11434/api/generate

Troubleshooting

You can check the Ollama troubleshooting guide for more information.