Adaline Providers

Adaline Providers

Concepts

Providers are configured LLM connections in an Adaline workspace. Models are provider-specific model records available to prompts and deployments.

Key terms:

• Provider — configured connection to a model provider
• providerId — provider record ID used to filter models
• provider — provider identifier such as openai, anthropic, or google
• Model — model record with model name and enabled flag

Configuration

Set ADALINE_API_KEY. Base URL: https://api.adaline.ai/v2.

Quick Triage

REST API

curl "https://api.adaline.ai/v2/providers" \
  -H "Authorization: Bearer $ADALINE_API_KEY"

curl "https://api.adaline.ai/v2/providers/provider_abc123?includeModels=true" \
  -H "Authorization: Bearer $ADALINE_API_KEY"

curl "https://api.adaline.ai/v2/models?providerId=provider_abc123" \
  -H "Authorization: Bearer $ADALINE_API_KEY"

SDK Usage

const providers = await adaline.providers.list();
const provider = await adaline.providers.get({ providerId, includeModels: true });
const models = await adaline.models.list({ providerId });

providers = await adaline.providers.list()
provider = await adaline.providers.get(provider_id=provider_id, include_models=True)
models = await adaline.models.list(provider_id=provider_id)

Best Practices

1. Discover provider IDs at startup instead of hardcoding them.
2. Filter models by providerId when building prompt/model selectors.
3. Use modelSettings only as capability metadata; pass actual runtime values in prompt/deployment config settings.
4. Cache provider/model lists briefly; they change less often than prompt traffic.

References

See references/api.md for schemas and examples.