Skip to main content

Dhenara Documentation

Build powerful AI applications with Dhenara's open-source frameworks

Dhenara AI

Open-source Python package for streamlined LLM integration across multiple providers

Unified API, Simplified

Access multiple AI providers through a consistent interface. Seamlessly switch between models without code changes using our lightweight yet powerful package.

Real-time Streaming

First-class support for streaming responses with accumulated results, making real-time applications simple to build.

Cost & Usage Tracking

Integrated token usage and cost tracking across all providers, with comprehensive analytics designed for both experimental and production applications.

Structured Output + Tools

Get typed, validated outputs (Pydantic) and provider-agnostic tool calling without rewriting your app for each provider.

Artifacts for Debugging

Capture normalized requests/responses and optional Python logs per call. Diagnose provider differences and validation failures fast.

Agent DSL (Deprecated)

Archived documentation for the legacy Dhenara Agent DSL (DAD).

Intuitive Agent DSL

Build sophisticated agent workflows by composing reusable components with a clean, programming language-like approach.

Comprehensive Observability

Free, Built-in Opentelemetry OpenTelemetry-based logging, tracing, and metrics collection for all agent activities, without subscriptions and payments.

Powerful Template Engine

Create sophisticated prompts and operations with dynamic content using variable substitution, conditional expressions, and hierarchical references.

Get Started Quickly

Dhenara makes it easy to build AI-powered applications. Start with Dhenara AI for provider-agnostic model interactions — then jump into the advanced recipes.

Agent DSL (Deprecated) — legacy docs for existing users.
pip install
# Recommended for most applications
pip install dhenara-ai

# Optional (deprecated): Agent DSL framework
pip install dhenara-agent