AI Gateway API Reference

Unified multimodal AI API with OpenAI-compatible LLM access.

Endpoints Overview

OpenAI-Compatible: Chat Completions

Endpoint: POST https://api.litestartup.com/client/v2/ai/chat/completions

Fully compatible with OpenAI's Chat Completions API. Use any OpenAI SDK by changing the base_url:

from openai import OpenAI

client = OpenAI(
    api_key="YOUR_API_KEY",
    base_url="https://api.litestartup.com/client/v2/ai"
)

response = client.chat.completions.create(
    model="deepseek-v3",
    messages=[{"role": "user", "content": "Hello!"}],
    temperature=0.7,
    max_tokens=4096
)

import OpenAI from 'openai';

const client = new OpenAI({
  apiKey: 'YOUR_API_KEY',
  baseURL: 'https://api.litestartup.com/client/v2/ai',
});

const response = await client.chat.completions.create({
  model: 'deepseek-v3',
  messages: [{ role: 'user', content: 'Hello!' }],
});

curl -X POST https://api.litestartup.com/client/v2/ai/chat/completions \
     -H 'Authorization: Bearer <your_api_key>' \
     -H 'Content-Type: application/json' \
     -d '{
  "model": "deepseek-v3",
  "messages": [
    {"role": "system", "content": "You are a helpful assistant."},
    {"role": "user", "content": "What is LiteStartup?"}
  ],
  "temperature": 0.7,
  "max_tokens": 4096
}'

Request Parameters

Response

{
  "id": "chatcmpl-a1b2c3d4e5f6",
  "object": "chat.completion",
  "created": 1721836800,
  "model": "deepseek-v3",
  "choices": [
    {
      "index": 0,
      "message": {"role": "assistant", "content": "LiteStartup is..."},
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 24,
    "completion_tokens": 128,
    "total_tokens": 152,
    "cost": "0.0012"
  }
}

Note: This endpoint only accepts LLM models. Non-LLM models (image, video, audio) will return a 400 error. Use /client/v2/ai/generations for those.

Unified: Generations

Endpoint: POST https://api.litestartup.com/client/v2/ai/generations

All modalities (LLM, image, video, audio, OCR) through a single endpoint. The model field determines which capability is invoked.

Headers

Request Body

Either messages (for LLM) or input (for other modalities) is required depending on the model type.

Examples by Modality

LLM Chat

curl -X POST https://api.litestartup.com/client/v2/ai/generations \
     -H 'Authorization: Bearer <your_api_key>' \
     -H 'Content-Type: application/json' \
     -d '{
  "model": "deepseek-v3",
  "messages": [
    {"role": "system", "content": "You are a helpful assistant."},
    {"role": "user", "content": "What is LiteStartup?"}
  ],
  "temperature": 0.7
}'

Response (OpenAI-compatible):

{
  "id": "chatcmpl-a1b2c3d4e5f6",
  "object": "chat.completion",
  "created": 1721836800,
  "model": "deepseek-v3",
  "choices": [
    {
      "index": 0,
      "message": {"role": "assistant", "content": "LiteStartup is..."},
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 24,
    "completion_tokens": 128,
    "total_tokens": 152,
    "cost": "0.0012"
  }
}

Image Generation (Sync)

curl -X POST https://api.litestartup.com/client/v2/ai/generations \
     -H 'Authorization: Bearer <your_api_key>' \
     -H 'Content-Type: application/json' \
     -d '{
  "model": "seedream-v5.0-lite-t2i",
  "input": {
    "prompt": "A futuristic city at sunset, cyberpunk style",
    "width": 1024,
    "height": 1024
  }
}'

Image OCR (Sync)

curl -X POST https://api.litestartup.com/client/v2/ai/generations \
     -H 'Authorization: Bearer <your_api_key>' \
     -H 'Content-Type: application/json' \
     -d '{
  "model": "ls-ocr",
  "input": {
    "image_url": "https://example.com/receipt.png"
  }
}'

Text to Video (Async)

curl -X POST https://api.litestartup.com/client/v2/ai/generations \
     -H 'Authorization: Bearer <your_api_key>' \
     -H 'Content-Type: application/json' \
     -d '{
  "model": "wan-2.7-t2v",
  "input": {
    "prompt": "A cat playing piano in a jazz bar",
    "duration": 5
  }
}'

Response (Async task submitted):

{
  "code": 200,
  "message": "Success",
  "data": {
    "task_id": 12345,
    "status": "processing",
    "model": "wan-2.7-t2v"
  },
  "meta": {
    "tool": "text-to-video",
    "provider": "atlascloud"
  }
}

Text to Speech (Sync)

curl -X POST https://api.litestartup.com/client/v2/ai/generations \
     -H 'Authorization: Bearer <your_api_key>' \
     -H 'Content-Type: application/json' \
     -d '{
  "model": "cosyvoice-2.0",
  "input": {
    "text": "Hello, welcome to LiteStartup!",
    "voice": "default"
  }
}'

Speech to Text / Transcription (Sync)

curl -X POST https://api.litestartup.com/client/v2/ai/generations \
     -H 'Authorization: Bearer <your_api_key>' \
     -H 'Content-Type: application/json' \
     -d '{
  "model": "sensevoice-v1",
  "input": {
    "audio_url": "https://example.com/meeting.mp3"
  }
}'

Response Formats

Sync Tool Response (Image/Audio/OCR)

{
  "code": 200,
  "message": "Success",
  "data": { ... },
  "meta": {
    "model": "model-id",
    "tool": "tool-type",
    "provider": "provider-name",
    "latency_ms": 1234
  }
}

Async Task Response (Video/Digital Human)

{
  "code": 200,
  "message": "Success",
  "data": {
    "task_id": 12345,
    "status": "processing",
    "model": "model-id"
  },
  "meta": {
    "tool": "tool-type",
    "provider": "provider-name"
  }
}

For async tasks, poll GET /client/v2/ai/tasks/{task_id} until status is completed:

curl https://api.litestartup.com/client/v2/ai/tasks/12345 \
     -H 'Authorization: Bearer <your_api_key>'

{
  "code": 200,
  "data": {
    "task_id": 12345,
    "status": "completed",
    "output": {
      "video_url": "https://cdn.litestartup.com/ai/output/..."
    }
  }
}

Task status values: processing, completed, failed.

Error Response

{
  "error": {
    "code": "model_not_found",
    "message": "The requested model 'xyz' is not available.",
    "type": "invalid_request_error"
  }
}

Error Codes