Documentation

Chat Completions

POST /v1/chat/completions is the OpenAI-compatible chat endpoint for SDKs, HTTP clients and tools that already speak Chat Completions.

Use it when

Your app already uses the OpenAI Chat Completions API shape.
You need standard messages, tools, tool_choice or stream behavior.
You want the broadest compatibility with OpenAI-style SDKs.

Code examples

curlJavaScriptPython

curl https://api.prismaticapi.com/v1/chat/completions \
  -H "Authorization: Bearer $PRISMATIC_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "claude-sonnet-4-6",
    "messages": [
      { "role": "system", "content": "You are a concise release assistant." },
      { "role": "user", "content": "Write three product taglines." }
    ],
    "stream": false,
    "max_tokens": 400
  }'

import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.PRISMATIC_API_KEY,
  baseURL: "https://api.prismaticapi.com/v1"
});

const completion = await client.chat.completions.create({
  model: "claude-sonnet-4-6",
  messages: [{ role: "user", content: "Write three product taglines." }],
  max_tokens: 400
});

import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ["PRISMATIC_API_KEY"],
    base_url="https://api.prismaticapi.com/v1",
)

completion = client.chat.completions.create(
    model="claude-sonnet-4-6",
    messages=[{"role": "user", "content": "Write three product taglines."}],
    max_tokens=400,
)

Parameters

model (string, required) — Public model ID returned by GET /v1/models.
messages (array, required) — Conversation turns with role and content.
max_tokens or max_completion_tokens (integer, optional) — Maximum generated output tokens.
stream (boolean, optional) — Enables OpenAI-style SSE streaming.
temperature, top_p, stop, tools and tool_choice are accepted when supported by the selected model.

Billing and limits

API-key and account rate limits are checked before generation.
Subscription quota is reserved before generation and settled afterward.
PAYG fallback can cover the request only when enabled and sufficiently funded.

PreviousAPI Endpoints NextResponses