OpenAI SDK Compatible

Use the official OpenAI SDK with TENSORAXIS and understand compatibility boundaries.

How It Works

TENSORAXIS provides a unified OpenAI-compatible entry point for common APIs. For Chat Completions, Responses, Embeddings, Images, Audio and similar endpoints, you usually only need to change two parameters when initializing the SDK:

api_key → your TENSORAXIS API Key
base_url → https://api.tensoraxis.com/v1

Some provider extensions and async task APIs have their own request-body conventions. In particular, video generation should not copy an upstream provider's raw content[] body directly; see Video Generation.

Python

pip install openai

from openai import OpenAI

client = OpenAI(
    api_key="your-tensoraxis-api-key",
    base_url="https://api.tensoraxis.com/v1",
)

Node.js / TypeScript

npm install openai

import OpenAI from 'openai'

const client = new OpenAI({
  apiKey: process.env.TENSORAXIS_API_KEY,
  baseURL: 'https://api.tensoraxis.com/v1',
})

Environment Variables (Recommended)

# .env
OPENAI_API_KEY=your-tensoraxis-api-key
OPENAI_BASE_URL=https://api.tensoraxis.com/v1

Once set, the OpenAI SDK reads these environment variables automatically:

from openai import OpenAI
client = OpenAI()

Supported Endpoints

Endpoint	Path	Description
Chat Completions	`POST /v1/chat/completions`	Chat completion with streaming support
Responses	`POST /v1/responses`	OpenAI Responses API
Embeddings	`POST /v1/embeddings`	Text vectorization
Models	`GET /v1/models`	List available models
Images	`POST /v1/images/generations`, `POST /v1/images/edits`	Image generation and editing, depending on model capability
Audio	`POST /v1/audio/transcriptions`, `POST /v1/audio/translations`, `POST /v1/audio/speech`	Audio transcription, translation and speech generation
Rerank	`POST /v1/rerank`	Reranking, typically called with direct HTTP
Video	`POST /v1/video/generations`, `GET /v1/video/generations/{task_id}`	Async video task API; see Video Generation
OpenAI/Sora Video	`POST /v1/videos`, `GET /v1/videos/{task_id}`	OpenAI/Sora-style video compatibility API

Video API Notes

Video generation is asynchronous. For Seedance / Doubao video models, TENSORAXIS accepts the unified task request:

{
  "model": "doubao-seedance-1-0-pro-250528",
  "prompt": "A cinematic cat watching rain by the window",
  "images": ["https://example.com/cat.png"],
  "metadata": {
    "resolution": "1080p",
    "ratio": "16:9",
    "duration": 5
  }
}

The upstream Volcengine Ark content[] structure is generated by the adapter. End users do not need to send it manually.