SDKs & Libraries

Overview

Fluxpool's API is OpenAI-compatible. The recommended integration path today is the official openai Python or Node.js SDK — verified end-to-end against production with the real npm package.

The official openai SDK works with Fluxpool for image generation and model listing. Video generation needs a raw HTTP request (see Video Generation) because the SDK doesn't have a video helper.

Installation

If you don't have the OpenAI SDK installed yet:

pip install openai

npm install openai

Requires openai >= 1.0.0 (Python) or openai >= 4.0.0 (Node.js).

Image Generation

from openai import OpenAI

client = OpenAI(
    base_url="https://api.fluxpool.ai/v1",  # ← Only change
    api_key="fp_live_xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx"
)

response = client.images.generate(
    model="flux-2-pro",
    prompt="A dragon perched on a neon-lit Tokyo rooftop, 8K",
    size="1024x1024"
)

# response.data[0].url is an S3 presigned URL — poll until ready
print(response.data[0].url)

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://api.fluxpool.ai/v1",  // ← Only change
  apiKey: "fp_live_xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx",
});

const response = await client.images.generate({
  model: "flux-2-pro",
  prompt: "A dragon perched on a neon-lit Tokyo rooftop, 8K",
  size: "1024x1024",
});

// response.data[0].url is an S3 presigned URL — poll until ready
console.log(response.data[0].url);

Replace the fp_live_... placeholder with your real key from API / MCP Access. See Authentication for details.

Listing Models

The OpenAI SDK's models.list() hits Fluxpool's GET /v1/models:

for m in client.models.list().data:
    print(m.id)

Video Generation

The OpenAI SDK does not have a videos helper. For video generation, POST directly to /v1/videos/generations:

import requests

response = requests.post(
    "https://api.fluxpool.ai/v1/videos/generations",
    headers={"Authorization": "Bearer fp_live_xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx"},
    json={
        "model": "seedance-v2",
        "prompt": "A drone shot flying through cherry blossom trees",
        "duration": 5,
        "resolution": "720p"
    }
).json()

gen_id = response["id"]
# Poll /v1/generations/{gen_id} or the presigned URL until ready.

Full polling example: see the Video Generation guide.

Native Fluxpool SDKs

REST API

No SDK needed. Any HTTP client works.

curl https://api.fluxpool.ai/v1/images/generations \
  -H "Authorization: Bearer fp_live_xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "flux-2-pro",
    "prompt": "Product photo of sneakers on marble, studio lighting",
    "size": "1024x1024"
  }'

Community libraries

No community-maintained Fluxpool packages are published yet. The OpenAI-compatible REST API works from any language that can speak HTTP — see the API Reference.

Building an integration? Email hi@fluxpool.ai and we'll link it here.