Multi-provider AI support: OpenAI, Anthropic, and Ollama (local)

[singret]

What changed

Regen now supports three AI providers out of the box. You can switch between them from Settings → System → AI Provider — no restart required.

Provider	What you need	Best for
OpenAI	API key	Cloud, highest quality
Anthropic	API key	Cloud, strong reasoning
Ollama	Running Ollama instance	Fully local, no data leaves your server

Switching providers

In the UI → Settings → System → AI Provider dropdown. Each provider shows only the fields it needs (key, model name, or base URL).

Via environment variables:

AI_PROVIDER=openai        # or anthropic, ollama
OPENAI_API_KEY=sk-...
ANTHROPIC_API_KEY=sk-ant-...
OLLAMA_BASE_URL=http://localhost:11434

Existing deployments with OPENAI_API_KEY set continue to work without any changes — AI_PROVIDER defaults to openai.

Running fully local with Ollama

Add to your docker-compose.yml:

services:
  ollama:
    image: ollama/ollama
    volumes:
      - ollama_data:/root/.ollama
    ports:
      - "11434:11434"

  regen:
    environment:
      AI_PROVIDER: ollama
      OLLAMA_BASE_URL: http://ollama:11434
      OLLAMA_MODEL: llama3.1:8b

volumes:
  ollama_data:

Pull the model once: docker exec <ollama-container> ollama pull llama3.1:8b

Recommended models: llama3.1:8b (minimum), llama3.1:70b (best quality).

See the README for the full Ollama setup guide.