# Dari

## Docs

- [API Keys](https://docs.dari.dev/api-keys.md): Create Routing and Management keys.
- [Router API Overview](https://docs.dari.dev/api-reference/overview.md): Send routed model traffic over HTTP.
- [Chat Completions API](https://docs.dari.dev/api-reference/router/chat-completions.md): Send a routed OpenAI-compatible chat completion request
- [Responses API](https://docs.dari.dev/api-reference/router/responses.md): Send a routed OpenAI Responses request
- [Authentication](https://docs.dari.dev/authentication.md): Choose the credential for each Dari workflow.
- [Router Evals](https://docs.dari.dev/evals/overview.md): Add benchmark scores as evidence for model selection.
- [Dari Auto Router](https://docs.dari.dev/framework/auto-router.md): Use Dari's hosted routing intelligence in your self-hosted router.
- [Custom Executors](https://docs.dari.dev/framework/custom-executors.md): Call models or transports outside the built-in Pi runtime.
- [Advanced Decision Pipeline](https://docs.dari.dev/framework/decision-pipeline.md): Prepare, inspect, and finalize deterministic routing decisions.
- [Models](https://docs.dari.dev/framework/models.md): Declare model capabilities and understand eligibility filtering.
- [Router Framework Overview](https://docs.dari.dev/framework/overview.md): Build and self-host an LLM router with pluggable policies and provider execution.
- [Pi Runtime And Credentials](https://docs.dari.dev/framework/pi-runtime.md): Execute selected models through Pi AI while keeping credentials in the host.
- [Policies](https://docs.dari.dev/framework/policies.md): Write routing policies that choose between eligible models.
- [Protocols And Streaming](https://docs.dari.dev/framework/protocols-and-streaming.md): The shared OpenAI and Anthropic request, response, and stream contract.
- [Framework Quickstart](https://docs.dari.dev/framework/quickstart.md): Run Router Core and serve your first routed request.
- [Speculative Routing](https://docs.dari.dev/framework/speculative-routing.md): Overlap provider execution with route selection in a custom host.
- [Dari Overview](https://docs.dari.dev/index.md): Use Dari's managed router or build and self-host your own.
- [Claude Code](https://docs.dari.dev/router/coding-agents/claude-code.md): Run Claude Code through a Dari Router.
- [Codex](https://docs.dari.dev/router/coding-agents/codex.md): Run Codex through a Dari Router.
- [Pi](https://docs.dari.dev/router/coding-agents/pi.md): Run Pi through a Dari Router.
- [Configure Models](https://docs.dari.dev/router/configure-models.md): Choose the models and reasoning levels a router may select.
- [Configure Provider Keys](https://docs.dari.dev/router/configure-provider-keys.md): Choose managed or BYOK credentials for router traffic.
- [Create A Router](https://docs.dari.dev/router/create-a-router.md): Create an endpoint that routes across your enabled models.
- [Model Activity](https://docs.dari.dev/router/model-activity.md): Read model usage, spend, latency, outcomes, and routing transitions from the CLI or Platform API.
- [Managed Router Overview](https://docs.dari.dev/router/overview.md): Route model requests through Dari's hosted platform.
- [Router Quickstart](https://docs.dari.dev/router/quickstart.md): Send your first request through Dari's default model router.
- [Send Chat Completions](https://docs.dari.dev/router/send-chat-completions.md): Send OpenAI-compatible requests through a router.
- [Speculative Routing](https://docs.dari.dev/router/speculative-routing.md): Reduce follow-up latency by overlapping provider execution with route selection.
- [Stream Chat Completions](https://docs.dari.dev/router/stream-responses.md): Stream routed output over Server-Sent Events.
- [Use Tool Calls](https://docs.dari.dev/router/use-tool-calls.md): Send OpenAI-style tools through a router.

## OpenAPI Specs

- [openapi](https://docs.dari.dev/api-reference/openapi.json)