Skip to main content

Quick Start with Dhenara AI

This guide gets you from installation to the first successful model call using the public ResourceConfig flow.

Setup​

First, make sure you have Dhenara installed:

pip install dhenara-ai

Then create a secret directory and place dai_credentials.yaml inside it. For a minimal Anthropic-only quick start:

mkdir -p /path/to/secrets
export DAI_SECRET_CONFIG_DIR=/path/to/secrets
/path/to/secrets/dai_credentials.yaml
anthropic:
  api_key: <YOUR_ANTHROPIC_API_KEY>

If you prefer, you can skip DAI_SECRET_CONFIG_DIR and pass an explicit credentials file path to ResourceConfig.load_from_file() instead.

Basic Text Generation​

from dhenara.ai import AIModelClient
from dhenara.ai.types import AIModelAPIProviderEnum, AIModelCallConfig, ResourceConfig

resource_config = ResourceConfig()
resource_config.load_from_file(credentials_file=None, init_endpoints=True)

endpoint = resource_config.get_model_endpoint(
    model_name="claude-haiku-4-5",
    api_provider=AIModelAPIProviderEnum.ANTHROPIC,
)
if endpoint is None:
    raise RuntimeError("No Anthropic endpoint configured for claude-haiku-4-5")

client = AIModelClient(
    model_endpoint=endpoint,
    config=AIModelCallConfig(
        max_output_tokens=1024,
        reasoning=False,
        streaming=False,
    ),
    is_async=False,
)

response = client.generate(
    prompt="What are three ways to improve productivity?",
    instructions=[
        "Be specific and actionable.",
    ],
)

if response.chat_response:
    print(response.chat_response.text())

load_from_file(credentials_file=None, init_endpoints=True) tells Dhenara AI to discover dai_credentials.yaml from DAI_SECRET_CONFIG_DIR and initialize all compatible foundation-model endpoints.

Enable reasoning (optional)​

If your chosen model supports reasoning/thinking, set reasoning=True.

client = AIModelClient(
    model_endpoint=endpoint,
    config=AIModelCallConfig(
        reasoning=True,
        reasoning_effort="medium",  # optional; normalized across providers
        max_reasoning_tokens=2000,  # optional; ignored by some providers
        max_output_tokens=1024,
    ),
    is_async=False,
)

response = client.generate(prompt="Explain the tradeoffs between TCP and UDP")

print("Answer:\n", response.chat_response.text())
print("Reasoning (if exposed by provider):\n", response.chat_response.reasoning())

Async usage​

import asyncio


async def main():
    async_client = AIModelClient(model_endpoint=endpoint, is_async=True)
    async with async_client as c:
        response = await c.generate_async(prompt="Give me 3 meal prep ideas")
        print(response.chat_response.text())


asyncio.run(main())

Next Steps​