Skip to content

Providers

Cristiano Carvalho edited this page Apr 2, 2026 · 10 revisions

Aludel supports cloud and local providers through a shared configuration model, while preserving provider-specific behavior in config. Each provider is a different reagent: same apparatus, different temperament.

image

Comparison

Provider Deployment API Key Best Fit Vision Support
OpenAI Cloud Required Fast setup and broad model coverage Yes, with PDF-to-image conversion
Anthropic Cloud Required Strong long-form and document workflows Yes, with native PDF support
Ollama Local Not required Local development and private evaluation Yes, via ImageMagick conversion

Configuration Shape

All providers follow the same top-level shape, which keeps the outer vessel familiar even when the contents change:

{
  "provider_type": "openai | anthropic | ollama",
  "model": "provider-specific-model",
  "api_endpoint": "optional override",
  "config": {}
}

Provider Examples

OpenAI

Requires the OPENAI_API_KEY environment variable.

{
  "provider_type": "openai",
  "model": "gpt-4o",
  "config": {
    "temperature": 0.7,
    "max_tokens": 1000
  }
}

Common models: gpt-4o, gpt-4o-mini, gpt-3.5-turbo, o1, o1-mini

Anthropic

Requires the ANTHROPIC_API_KEY environment variable.

{
  "provider_type": "anthropic",
  "model": "claude-3-5-sonnet-20241022",
  "config": {
    "temperature": 1.0,
    "max_tokens": 4096
  }
}

Common models: claude-3-5-sonnet-20241022, claude-3-5-haiku-20241022, claude-3-opus-20240229

Ollama

No API key is required. Ollama runs locally and is well suited to private experiments and fast iteration.

ollama pull llama3.1:8b
{
  "provider_type": "ollama",
  "model": "llama3.1:8b",
  "api_endpoint": "http://localhost:11434/api/chat",
  "config": {
    "temperature": 0.8,
    "num_ctx": 4096
  }
}

Common Parameters

Parameter Type Default Description
temperature float 0.7 Randomness (0.0-2.0)
max_tokens integer 1000 Max output length
top_p float 1.0 Nucleus sampling

Additional provider-specific parameters:

  • OpenAI: frequency_penalty, presence_penalty, seed
  • Anthropic: top_k
  • Ollama: num_ctx, num_predict, repeat_penalty

API Keys

Keys are read from the environment and are not stored in the database:

export OPENAI_API_KEY=sk-...
export ANTHROPIC_API_KEY=sk-ant-...

Cost Tracking

Cost is estimated automatically from token counts and provider pricing data. These values are useful for comparison and trend analysis, but they should be treated as operational estimates rather than billing truth.

API Reference

create_provider(attrs)
update_provider(provider, attrs)
delete_provider(provider)
get_provider!(id)
list_providers()
list_enabled_providers()
list_by_type(type)

Clone this wiki locally