Provider Setup

Pick a provider based on privacy, cost, and latency. Local providers keep data on your machine. Cloud providers are convenient but require API keys.

Quick Start

1. Choose a provider.
2. Add credentials if needed.
3. Create a profile.
4. Run victor chat -p <profile>.

Local Example (Ollama)

# Start server
ollama serve

# Pull a model
ollama pull <model-id>

profiles:
  local:
    provider: ollama
    model: <model-id>

providers:
  ollama:
    base_url: http://localhost:11434

Cloud Example (OpenAI)

export OPENAI_API_KEY="sk-..."

profiles:
  cloud:
    provider: openai
    model: <model-id>

providers:
  openai:
    api_key: ${OPENAI_API_KEY}

Other Providers

Local:
- Ollama
- LM Studio
- vLLM
- llama.cpp

Cloud:
- Anthropic
- OpenAI
- Google
- xAI
- Mistral
- DeepSeek
- Together
- Fireworks
- OpenRouter
- Groq
- Moonshot
- Cerebras

Enterprise:
- Azure OpenAI
- AWS Bedrock
- Google Vertex AI

Platforms:
- Hugging Face
- Replicate

Tips

• Use ../API_KEYS_CONFIGURATION.md for credential options.
• If tool calling is inconsistent, try another model or switch to keyword tool selection.
• For offline use, combine local models with ../embeddings/AIRGAPPED.md.