You can connect Codelantis to an AI model provider (e.g. OpenAI) and perform code reviews using an AI model (LLM) via the respective API. Note that this may lead to additional costs by the AI vendor.
When an AI provider has been configured, users can request to review a file of a pull request (merge request) with the push of the button above each file (shortcut: Ctrl + Shift + R).
Codelantis will send the file contents (or parts of it) to the configured AI provider and ask for review comments.
Review comments raised by the LLM will then be shown directly in the code. Unlike human review comments, these comments are not pushed to GitHub / GitLab. The AI comments can either be resolved (when the issue has been fixed) or deleted (if irrelevant suggestion were being made).
Currently, we support the following AI providers:
| Provider | Name |
| OpenAI | GPT models (e.g. gpt-4o) |
Gemini (e.g. gemini-2.5-pro-preview-06-05) |
To configure an AI provider in Codelantis, you'll need to set the following environment variables:
AI_PROVIDER: The name of the provider: google or openaiAI_MODEL_NAME: The model to use, e.g. gpt-4oAI_ACCESS_TOKEN: The API access token by the respective providerAfter these environment variables have been set, the AI review functionality will be available to all users.