Dhee AI is an adaptive AI orchestration platform and unified API. It automatically detects request intent, selects and composes verified domain skills, attaches sandboxed tools, selects the optimal Dheeman cognitive model, routes inference across AI providers, and validates output through a single endpoint.

How does Dhee differ from an AI model gateway?

A basic model gateway only proxies requests to cheaper or faster LLM providers. Dhee orchestrates the full execution stack: determining required domain expertise (Skills), binding sandboxed capabilities (Tools), allocating intelligence depth (Dheeman Models), and auditing output schemas before returning.

What is Dheeman Auto?

Dheeman Auto is the flagship meta-routing engine. When calling model: "dheeman-auto", Dhee evaluates task requirements, context size, and performance constraints in real time to select the optimal model and provider route automatically.

Is Dhee OpenAI-compatible?

Yes. Dhee provides drop-in compatibility with the OpenAI SDK. Developers point their client to baseURL "https://ai.dheelab.com/v1" with standard chat completions, streaming SSE, and tool calling supported immediately.

What happens when an underlying AI provider fails?

Dhee continuously monitors provider latency, health, and error rates. If an upstream provider returns 429 rate limits or 5xx outages, Dhee re-routes execution instantly to healthy standby providers without dropping client connections.

ADAPTIVE AI ORCHESTRATION

One request. The right skills, tools and intelligence.

Dhee understands every request, assembles the right verified expert skills, selects the tools and Dheeman model, routes across AI providers, and validates the result — through one API.

Start Building Free See How Dhee Works

✓ One API

✓ Adaptive Skills

✓ Multi-model

✓ Automatic Fallback

HOW DHEE WORKS

From intent to validated outcome.

Instead of sending raw prompts to a single AI model, Dhee orchestrates the full execution stack.

Intent Understanding

Deconstructs complex user prompts to identify the real task, required domain competencies, and output constraints.

Skill Composition

Selects verified expert skills from the catalog and combines multiple skills when a cross-discipline task requires them.

Guarded Tool Execution

Determines which external tools are needed (web search, data extraction, code execution) and binds them safely under strict egress policies.

Virtual Model Selection

Selects the right Dheeman intelligence depth (Flash, Pro, Reasoning, Coding) based on task complexity and service tier.

Resilient Multi-Provider Routing

Routes across AI providers (AWS Bedrock, Azure OpenAI, GCP Vertex, Anthropic) with real-time health checks and automatic failover.

Strict Output Validation

Verifies the response against output contracts, schema definitions, and quality metrics before returning to the caller.

Transparent Economics

Simple platform plans. Transparent token rates.

Choose a subscription tier for platform throughput and included usage credits, then pay exact retail rates for model usage.

Free

$0 /month

Includes $0 usage credit each month.

✓ 5 RPM
✓ 25k TPM
✓ Standard Routing

Start Free

Starter

$9 /month

Includes $3 usage credit each month.

✓ 30 RPM
✓ 200k TPM
✓ Standard Routing

Pro

$29 /month

Includes $10 usage credit each month.

✓ 120 RPM
✓ 1,000k TPM
✓ Priority Routing

Business

$99 /month

Includes $40 usage credit each month.

✓ 300 RPM
✓ 3,000k TPM
✓ Priority Routing

FREQUENTLY ASKED QUESTIONS

Common questions about Dhee.

What is Dhee?

Dhee is an Adaptive AI Orchestration Layer. When you send a request, Dhee automatically understands your intended outcome, selects verified expert Skills, binds required Tools, selects the optimal Dheeman intelligence model, routes across AI providers, and validates the result — returning everything through one unified API.

How is Dhee different from a model gateway?

A model gateway simply forwards prompts to an LLM provider based on price or latency. Dhee is an orchestration intelligence: it decouples domain expertise from raw models, automatically assembling the workflows, tools, and cognitive depth required to solve multi-step business objectives.

What are Dhee Skills?

Dhee Skills are verified domain workflows (such as SEO Audits, Financial Anomaly Detection, or Code Review) that define how specialized tasks should be executed. They provide step-by-step expertise so the model doesn’t rely on superficial prompt guesswork.

Deploy intelligent workflows in minutes.

One API. Adaptive skill orchestration. Multi-model resilience. Start free and scale seamlessly.

Start Building Free Explore Docs