Install Jentic One Beta
Jentic One is a self-hosted execution layer for AI agents. It lets your agent call the RouterPlex API, or any other public or private API you need. You set the rules, the agent never sees your credentials, and every call is logged.
Two steps, two machines. Install the instance in a safe environment, then register your agent from wherever it runs.
Step 1: Jentic One Host machine
# On the machine that will host your Jentic One instance:
curl -fsSL "https://jentic.com/install.sh?src=apis&api=%2Fapis%2Frouterplex.com%2Frouterplex" | shStep 2: Agent machine
# On the machine where your agent runs (keep this separate from the instance):
curl -fsSL "https://jentic.com/install.sh?src=apis&api=%2Fapis%2Frouterplex.com%2Frouterplex" | sh
jentic register # connects your agent to your Jentic One instanceJentic One is in public beta. The setup above keeps your agent separate from the instance, which is what you want before using real credentials: an agent running as the same OS user as Jentic One can read its stored keys directly. Just evaluating? A single local install is fine to start. See the secure deployment guide for the tiers.
What an agent can do with RouterPlex API.
Send OpenAI-compatible chat completions with a model ID and a messages array, optionally streamed
Send Anthropic-compatible messages with a model, messages, and an optional max_tokens cap
Generate images from a prompt with a chosen model, size, and quality
Serve Codex custom model providers through the Responses-format route with model and input
GET STARTED
List the model IDs available to the authenticated RouterPlex API key
Patterns agents use RouterPlex API for, with concrete tasks.
★ AI agent reaching OpenAI-style and Anthropic-style models with one key
An AI agent that alternates between OpenAI-format and Anthropic-format requests needs a single credential and base URL. Through Jentic the agent searches for 'send a chat completion', loads the createChatCompletion operation with its input schema, and calls it with the RouterPlex key injected at execution time. For Anthropic-style calls it uses createMessage on the same gateway, choosing a model ID returned by listModels.
Call GET /models to find an available model ID, then POST /chat/completions with that model and a messages array containing the user's question
Codex custom model provider
A developer pointing Codex at a custom model provider needs a Responses-format endpoint rather than Chat Completions. RouterPlex exposes POST /responses for exactly that client, taking a model ID and an input value, with optional streaming. Other OpenAI-compatible clients are directed to Chat Completions unless they require the Responses shape.
Send POST /responses with a model ID from GET /models and an input prompt, with stream set to false, and return the response
Image generation through the gateway
An app that produces illustrations wants image generation on the same key as its text models. RouterPlex accepts POST /images/generations with a model such as gpt-image-2, a prompt, and optional size and quality fields. The app keeps one credential and one base URL for both text and image work.
Call POST /images/generations with model gpt-image-2, a product description as the prompt, and a size value, then return the generated image
5 endpoints — the routerplex api forwards model requests through an ai gateway that speaks both openai-compatible and anthropic-compatible formats behind one routerplex api key.
METHOD
PATH
DESCRIPTION
/chat/completions
Create an OpenAI-compatible chat completion
/messages
Create an Anthropic-compatible message
/responses
Create a Responses-format response for Codex
/images/generations
Generate an image
/models
List models available to the key
/chat/completions
Create an OpenAI-compatible chat completion
/messages
Create an Anthropic-compatible message
/responses
Create a Responses-format response for Codex
/images/generations
Generate an image
/models
List models available to the key
What agents get from Jentic-routed access to this vendor.
Setup
Wiring RouterPlex by hand means setting the /v1 base URL, attaching the bearer key to each request, and keeping OpenAI-format and Anthropic-format request bodies straight across routes. Through Jentic you install once, import RouterPlex from the API Directory, store the key once, and your agent calls it.
Permission scoping
RouterPlex splits its surface into five paths, from /chat/completions to /images/generations, so a rule can bound your agent to exactly the routes a task needs. You choose the operations it may call, so image generation or the Codex Responses route is not included unless you add it.
Credential isolation
Your RouterPlex 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.
Intent-based discovery
Agents search Jentic by intent such as 'send a chat completion' or 'generate an image', and Jentic returns the matching RouterPlex operation with its input schema, including required fields like model and messages, so the agent calls it correctly.
Alternatives and complements available in the Jentic catalogue.
Specific to using RouterPlex API through Jentic.
What authentication does the RouterPlex API use?
Per its OpenAPI spec, the RouterPlex API uses HTTP bearer authentication with a RouterPlex API key, sent in the Authorization header on all five operations. You create and manage keys in the RouterPlex dashboard. Through Jentic the key is held by your own instance and attached at call time, so the agent never handles it.
Can I send Anthropic-style requests with the RouterPlex API?
Yes. POST /messages is an Anthropic-compatible route that takes a model ID and a messages array, with optional max_tokens and stream fields. It runs on the same base URL and key as the OpenAI-compatible POST /chat/completions route.
What are the rate limits for the RouterPlex API?
The OpenAPI spec does not specify rate limits. Check the RouterPlex dashboard and documentation at https://routerplex.com for the current limits on your key.
Is there a RouterPlex MCP server?
You do not need an MCP server to give your agent RouterPlex. Jentic connects it directly from the API Directory: import it, store your RouterPlex API key once, and your agent calls the chat, messages, image, and model operations on demand without extra tool definitions in its context.
How do I pick a model and send a chat completion through Jentic?
Search Jentic for 'list available models', run GET /models to get the IDs your key can use, then search for 'send a chat completion' and execute POST /chat/completions with one of those IDs and your messages. To run it on your own infrastructure, install Jentic One from its GitHub repo.
Can I limit what my agent is allowed to do with the RouterPlex API?
Yes. Write a rule that allows only POST /chat/completions and GET /models, so the agent can chat with the models on your key but cannot generate images or use the Responses route. Add other routes when a task needs them, and every call the agent makes is logged.
For Agents
Send OpenAI-compatible chat completions, Anthropic-compatible messages, Codex Responses calls, and image generations through one RouterPlex key, and list the models that key can use.
Use for: List all models my RouterPlex key can use, Send a chat completion through RouterPlex, I want to send an Anthropic-style message with a token cap, Generate an image from a prompt with a specific size
Not supported: Does not create embeddings, transcribe audio, or fine-tune models. Use for routing chat, message, Codex Responses, and image generation requests only.
The RouterPlex API forwards model requests through an AI gateway that speaks both OpenAI-compatible and Anthropic-compatible formats behind one RouterPlex API key. It accepts chat completions, Anthropic-style messages with an optional max_tokens cap, and image generation requests with a model, prompt, size, and quality. A Responses route serves Codex custom model providers, and a models listing returns the model IDs available to the authenticated key.
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.
Base layer of spec validity and structural soundness.
Aggregated quality score from linter diagnostics, weighted by severity.
Percentage of `$ref` references that resolve successfully.
Checks whether the API description parses successfully and conforms to its declared specification (e.g., OpenAPI).
Structural correctness score based on schema issues using logarithmic dampening.
Clarity, completeness, and ingestion readiness for developers and tooling.
How richly the API is illustrated with examples.
Percentage of examples that conform to their schemas.
Percentage of operations with complete response definitions (success, client error, server error).
Health of API ingestion, bundling, and resolution within Jentic pipelines.
Semantic breadth, depth, and agent comprehension for AI systems.
Coverage of descriptions across API elements.
Coverage of RFC 9457 Problem Details for error responses.
Coverage, uniqueness, and casing consistency of operationIds for AI inference.
Coverage of summaries across operations/tags/info.
Functional utility, complexity comfort, and AI orchestration readiness.
Agent comfort level based on API operational and structural complexity.
Trust, risk posture, and security compliance.
Average quality of security schemes based on authentication method strength (weakest link for OAuth2).
Findability, semantic richness, and reasoning readiness.
Clarity and depth of descriptions across API elements.
Score it yourself
Every API in the directory is allowlisted, so you can re-score it with no key required.
npx @jentic/api-scorecard-cli score <openapi-url>