canonical: https://jentic.com/apis/getorquesta.com/getorquesta

# Getorquesta Orquesta Public API

The Orquesta Public API centres on Batuta, an OpenAI-compatible chat-completions proxy that routes a request across multiple upstream models using per-prompt classification, so an application calls one endpoint and Orquesta picks the model. Alongside the proxy it exposes the organization's prompts, which can be listed, read with their context, submitted for execution, updated, and cancelled, plus the execution logs those runs produce. It also gives read access to projects, registered agents and their online status, and the users in scope.

## For AI agents

Route OpenAI-compatible chat completions across models through Orquesta's Batuta proxy, submit and cancel prompts, and read execution logs. Authenticates with a bearer token or an API key.

## Scope

Does not train or host the underlying models, store vector embeddings, or manage billing. Use for routing chat completions, prompt execution, and reading execution logs in Orquesta only.

## Capabilities

- Route chat completions across upstream models through the OpenAI-compatible Batuta proxy
- List the models available through the Batuta proxy
- Submit a prompt for execution and cancel a pending or running one
- List and read prompts in your organization with their context
- List execution logs filtered by prompt or project
- Read projects, registered agents and their online status, and users in scope

## Use cases

### Model-Routed Chat Completions

An AI agent connected through Jentic sends an OpenAI-compatible chat-completions request to Orquesta's Batuta proxy, which classifies the prompt and routes it to a suitable upstream model, so the agent gets a completion without picking or wiring a model itself. Calling one endpoint keeps the agent's integration stable while Orquesta decides where the request goes.

Example prompt: Send a chat-completions request to the Batuta proxy and return the routed model's completion

### Prompt Execution and Cancellation

A workflow submits a prompt for execution and cancels it if it is still pending or running, using the prompt operations to control long-running work. Reading the prompt back with its context lets the agent confirm what was run before acting on the result.

Example prompt: Submit this prompt for execution, then cancel it if it has not completed within the time limit

### Execution Log Review

A monitoring workflow lists execution logs by prompt or project to review what ran and how, pulling the record Orquesta keeps for each execution. This gives an agent or an operator a read-only trail without opening the dashboard.

Example prompt: List the execution logs for this project and summarise the most recent runs

## Key endpoints

| Method | Path | Description |
| --- | --- | --- |
| POST | `/api/v1/chat/completions` | Create a chat completion (OpenAI-compatible) |
| GET | `/api/v1/models` | List models available through Batuta |
| POST | `/api/v1/prompts` | Submit a prompt for execution |
| GET | `/api/v1/prompts` | List prompts in your organization |
| POST | `/api/v1/prompts/{id}/cancel` | Cancel a pending or running prompt |
| GET | `/api/v1/logs` | List execution logs by prompt or project |
| GET | `/api/v1/projects` | List projects in your organization |
| GET | `/api/v1/agents` | List registered agents and their status |

## Key resources

- **Batuta Proxy** — OpenAI-compatible chat completions routed across upstream models, plus the model list
- **Prompts** — Prompts listed, read with context, submitted for execution, updated, and cancelled
- **Logs** — Execution logs listed by prompt or project
- **Projects, Agents, and Users** — Read access to projects, registered agents with online status, and users in scope

## AI readiness

This API is usable in Jentic One now. Its AI-readiness score against Jentic's framework shows where it stands today and where improvements would make it even easier for agents to use.

- **Score:** 61 / 100
- **Maturity:** AI-Aware
- **Dimensions:**
  - Foundational Compliance: 54 / 100
  - Developer Experience & Jentic Compatibility: 65 / 100
  - AI-Readiness & Agent Experience: 48 / 100
  - Agent Usability: 94 / 100
  - Security: 55 / 100
  - AI Discoverability: 83 / 100
- **View full report:** https://jentic.com/apis/getorquesta.com/getorquesta/scorecard
- **How the score is calculated:** https://docs.jentic.com/reference/api-readiness-framework/overview/
- **More about the dimensions:** https://docs.jentic.com/reference/api-readiness-framework/specification/#dimensional-model-overview

### Score it yourself

Every API in the directory is allowlisted, so you can re-score it with no key required.

- **Score your own API:** https://jentic.com/scorecard.md
- **Scoring CLI agent skill:** https://github.com/jentic/jentic-api-scorecard/blob/main/skills/jentic-api-scorecard/SKILL.md

```sh
npx @jentic/api-scorecard-cli score <openapi-url>
```

## Why Jentic

- **Setup:** Wiring the Orquesta API by hand means choosing between its bearer token and X-API-Key credentials, carrying one on each request, and integrating the OpenAI-compatible proxy alongside the prompt and log endpoints yourself. Through Jentic you install once, import Orquesta from the API Directory, store the credential once, and your agent calls it.
- **Permission scoping:** Most of the Orquesta surface is read-only, so a rule can allow your agent to list prompts, logs, models, and projects while leaving the write operations out. You choose which operations it may call, so submitting or cancelling a prompt is not included unless you add it.
- **Credential handling:** Your Orquesta bearer token or API key is stored once, encrypted, by your own Jentic One instance and injected at execution time. It never enters the agent's prompt, logs, or context.
- **Discovery method:** Agents search Jentic by intent such as 'route a chat completion' or 'list execution logs', and Jentic returns the matching Orquesta operation with its input schema so the agent calls the right endpoint without reading the reference docs.

## Related APIs

- **OpenRouter API** — OpenAI-compatible router that sends requests across many upstream models
- **Portkey API** — AI gateway that routes model requests and adds observability and guardrails
- **Humanloop API** — Prompt management and evaluation platform for LLM applications
- **Langfuse API** — LLM observability that captures and traces the executions you route

## FAQ

### Is there an Orquesta MCP server?

You don't need an MCP server to give your agent the Orquesta API. Jentic connects it directly from the API Directory: import it, store your credential once, and your agent can route chat completions and read prompts straight away, with no extra server to run and no tool definitions to load into the agent's context.

### Can I limit what my agent is allowed to do with the Orquesta API?

Yes. Write a rule that allows only the read operations the agent needs, such as listing prompts, logs, and models, so it can review activity without submitting or cancelling prompts. Write operations like submitting or cancelling a prompt are not included unless you add them, and every call the agent makes is logged.

### What authentication does the Orquesta API use?

The Orquesta API accepts two credentials per its OpenAPI spec: a bearer token in the Authorization header, or an API key in the `X-API-Key` header. Through Jentic the credential is stored encrypted by your own instance and injected at call time, so it never appears in the agent's prompt or logs.

### How does the Batuta proxy choose a model?

The chat-completions endpoint is OpenAI-compatible and routes each request across multiple upstream models using per-prompt classification, so your agent calls one endpoint and Orquesta selects the model. You can list the models available through the proxy to see what it can route to.

### What are the rate limits for the Orquesta API?

The OpenAPI spec does not specify rate limits for the Orquesta API. For current limits and token formats, see the Orquesta documentation at https://getorquesta.com.

### How do I route a chat completion through Jentic?

Import the Orquesta API from the Jentic API Directory, then have your agent issue a request such as 'route this chat completion across models'. Jentic matches the intent to the chat-completions operation and returns its input schema so the agent sends the request correctly, with your stored credential injected at call time. To run it on your own infrastructure, install Jentic One from its GitHub repo.