Skip to content
Docs navigation: Open models / Overview

Open models

RightNow AIOpen models Updated

On this page

The open models API serves hosted open-weight models through an OpenAI-compatible Chat Completions API and an Anthropic-compatible Messages API. One workspace API key reaches them all, and credits pay for usage.

two protocols, one key
Text
Illustration of one workspace key reaching the open models through two protocols.

Two protocols, chat completions at /v1/chat/completions and Messages at /v1/messages, take the same workspace key. Both reach the open models on one host. Usage is paid from one credit balance, drawn as a bar.

The RightNow decision model has its own docs.

Call a hosted model

Set the base URL for your route, use a workspace API key and pass a model id.

The example calls glm-5-3; any available id from Open models works in its place. Read current capabilities, limits and prices from GET /v1/models.

TypeScript
import OpenAI from "openai";

// Server side only: never expose this key in browser code.
const client = new OpenAI({
  baseURL: "https://api.rightnowai.co/v1",
  apiKey: process.env.RIGHTNOW_API_KEY,
});
const response = await client.chat.completions.create({
  model: "glm-5-3",
  messages: [{ role: "user", content: "Write a TypeScript retry helper." }],
  max_tokens: 16384,
  reasoning_effort: "low",
});
console.log(response.choices[0]?.message?.content);

GET /v1/models is the authoritative list of hosted chat models. It also reports when one is temporarily paused.

Some names in requests and replies start with runinfra, from the platform's earlier name. They stay as they are.

Which route

RoutePathClient base URLRead
Chat completionsPOST /v1/chat/completionshttps://api.rightnowai.co/v1Chat completions and Responses
ResponsesPOST /v1/responseshttps://api.rightnowai.co/v1Chat completions and Responses
MessagesPOST /v1/messageshttps://api.rightnowai.coMessages

Read in this order

  1. Call an open model

    Get a workspace key, discover an available model and send a server-side chat request.

  2. Models

    Discover the model ids a workspace key can reach, and the limits the gateway enforces on them.

  3. Chat completions and Responses

    Chat and Responses request fields, tool loops, streamed events, usage, validation and output budgets.

  4. Streaming

    Stream chat completion deltas over Server-Sent Events, and ask for the usage frame.

  5. Errors

    The error envelope, every status and code, and whether retrying can help.

  6. Retries

    Retry a call, streaming or not, without starting duplicate inference work or a second charge.

  7. Limits

    What each purchase unlocks, what each 429 means, and how to back off.

  8. Sessions

    Keep a session's requests together so its prefix cache stays warm across turns and fan-out workers.

Pricing, access and help