Documentation

API Endpoints

The public API exposes model discovery plus OpenAI-compatible, Gemini-compatible and Anthropic-compatible generation routes.

Core text generation

POST/v1/chat/completions

Chat Completions

OpenAI-compatible chat completions with optional streaming.

POST/v1/responses

Responses

OpenAI Responses-compatible input, instructions and tools through the same billing layer.

POST/v1/models/{model}:generateContent

Gemini generateContent v1

Gemini-compatible non-streaming generation. The {model} path segment is the Prismatic public model ID.

POST/v1beta/models/{model}:generateContent

Gemini generateContent v1beta

Beta-version Gemini-compatible non-streaming generation for clients that target the v1beta Gemini surface.

POST/v1/models/{model}:streamGenerateContent

Gemini streamGenerateContent v1

Gemini-compatible SSE streaming. Do not set stream: true on generateContent; use this endpoint instead.

POST/v1beta/models/{model}:streamGenerateContent

Gemini streamGenerateContent v1beta

Beta-version Gemini-compatible SSE streaming for clients that target the v1beta Gemini surface.

POST/v1/images/generations

Image generations

OpenAI-compatible image generation. Prismatic requires model so routing, plan access and billing are explicit.

POST/v1/videos

Video generations

OpenAI-compatible video generation. Prismatic requires model and routes only to providers marked with video support.

POST/v1/messages

Messages

Anthropic-compatible Messages endpoint for Claude-native clients and coding agents.

Models and token estimates

GET/v1/models

List models

Returns enabled public model IDs and metadata in the shape expected by the request protocol. A single canonical model can appear under multiple public IDs when aliases are configured.

GET/v1/models/{model_id}

Get model

Returns one enabled model by public ID or model_not_found when unavailable.

POST/v1/messages/count_tokens

Count tokens

Estimates input tokens for an Anthropic-style Messages payload before generation.

Account

Read the authenticated account's quota and wallet balance with the same API key used for generation.

GET/v1/account/quota

Account quota

Returns the account's subscription quota windows (5h / 7d rolling, or the daily window for daily-mode plans) with limit / used / remaining and reset times, plus the PAYG wallet balance in USD. plan, windows and daily are null when the account has no active subscription.

Bash

curl https://api.prismaticapi.com/v1/account/quota \
  -H "Authorization: Bearer $PRISMATIC_API_KEY"

Image generation example

Use an image-capable public model ID from GET /v1/models. The request consumes subscription quota or PAYG through the same billing checks as text generation.

BashJavaScript

curl https://api.prismaticapi.com/v1/images/generations   -H "Authorization: Bearer $PRISMATIC_API_KEY"   -H "Content-Type: application/json"   -d '{
    "model": "imagen-4",
    "prompt": "A clean product mockup on a neutral desk",
    "image_url": ["https://example.com/reference.png"],
    "negative_prompt": "clutter, text artifacts",
    "aspect_ratio": "16:9",
    "quality": "high",
    "style": "natural",
    "seed": 42,
    "response_format": "b64_json",
    "n": 1
  }'

import OpenAI from "openai";

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

const image = await client.images.generate({
  model: "imagen-4",
  prompt: "A clean product mockup on a neutral desk",
  image_url: "https://example.com/reference.png",
  negative_prompt: "clutter, text artifacts",
  aspect_ratio: "16:9",
  quality: "high",
  style: "natural",
  seed: 42,
  response_format: "b64_json",
  n: 1
});

Other surfaces

Audio, embeddings and moderation endpoints are not part of Prismatic v1 yet.
Do not configure clients to call unsupported routes until they are exposed by the API.
The dashboard remains the source of truth for account usage and billing views.

PreviousAuthentication NextChat Completions