ollama-openai-proxy is a proxy server that emulates the REST API of ollama/ollama. Requests are forwarded to OpenAI. The connection to OpenAI is made via a third-party sashabaranov/go-openai. This allows you to use OpenAI models via a proxy from Jetbrains AI Assistant.

# Create .env file and fill OPENAI_API_KEY variable.
cp -p .env.local .env

docker compose up --build -d

1. Open Settings.
2. Move Settings > Tools > AI Assistant > Models.
3. Check Enable Ollama. Check whether the Test Connection is successful.
4. Select the model you want to use under Core features and Instant helpers.
5. Check Offline mode.
6. Press Apply, then OK.
7. When you open AI Chat, you will be able to select a model from Ollama.