AI/ML API Node.js SDK

The official TypeScript/JavaScript SDK for AI/ML API - access 400+ AI models (GPT-4, Claude, Gemini, DeepSeek, etc.) with an OpenAI-compatible interface.

Installation

``bash
npm install @ai-ml.api/aimlapi-sdk-node
`

Quickstart

$3

Create a .env file with your API key:

`bash
AIML_API_KEY=your-api-key-here
`

$3

`typescript
import { AIMLAPI } from "@ai-ml.api/aimlapi-sdk-node";
import dotenv from "dotenv";

dotenv.config();

const client = new AIMLAPI();
`

$3

Use npx tsx to run TypeScript/JavaScript files directly:

`bash
npx tsx your-script.ts
`

Or with environment file:

`bash
AIML_API_KEY=your-key npx tsx your-script.ts
`

Available Resources

- Chat & Messages: OpenAI-compatible chat, Anthropic messages
- Images & Vision: Image generation, OCR, vision analysis
- Video & Audio: Video generation, TTS, speech-to-text
- Music & Search: Music generation, Bagoodex search
- Utilities: Batches, embeddings, account management

Examples

$3

`typescript
import { AIMLAPI } from "@ai-ml.api/aimlapi-sdk-node";
import dotenv from "dotenv";

dotenv.config();

const client = new AIMLAPI();

const completion = await client.chat.completions.create({
model: "gpt-4o",
messages: [
{ role: "system", content: "You are a helpful assistant." },
{ role: "user", content: "Explain quantum computing." },
],
});

console.log(completion.choices[0]?.message?.content);
`

$3

`typescript
const stream = await client.chat.completions.create({
model: "gpt-4o",
messages: [{ role: "user", content: "Write a story." }],
stream: true,
});

for await (const chunk of stream) {
process.stdout.write(chunk.choices[0]?.delta?.content ?? "");
}
`

$3

Long-running video generation with automatic status polling:

`typescript
// Generate video and wait for completion
const result = await client.video.createWithPolling(
{
model: "openai/sora-2-t2v",
prompt: "A menacing evil dragon appears above a mountain",
resolution: "720p",
aspect_ratio: "16:9",
duration: 4,
},
20, // maxAttempts
5000 // pollIntervalMs
);

console.log("Video status:", result.status);
console.log("Video URL:", result.video?.url);
`

Alternative: Generate first, then poll separately:

`typescript
// Start generation
const generation = await client.video.generate({
model: "openai/sora-2-t2v",
prompt: "A robot dancing",
resolution: "720p",
aspect_ratio: "16:9",
});

// Poll for status
const status = await client.video.getStatus(generation.id);
console.log("Status:", status.status);
`

$3

Music generation supports multiple models (ElevenLabs, Minimax, Google Lyria):

`typescript
// Generate music and wait for completion
const music = await client.music.createWithPolling(
{
model: "elevenlabs/eleven_music",
prompt: "Lo-fi hip-hop ambient music",
music_length_ms: 20000,
},
30, // maxAttempts
120 // timeoutSeconds
);

console.log("Music status:", music.status);
console.log("Audio URL:", music.audio_file?.url);
`

Using Minimax Music-2.0 with lyrics:

`typescript
const music = await client.music.generateMusic({
model: "minimax/music-2.0",
prompt: "Electronic dance music",
lyrics: "[Verse]\nElectronic beats\n[Chorus]\nDance all night",
audio_setting: {
format: "mp3",
sample_rate: 44100,
},
});

console.log("Music ID:", music.id);
`

$3

Transcribe audio with automatic polling for completion:

`typescript
// Transcribe audio with polling
const result = await client.speechToText.createWithPolling(
{
model: "aai/slam-1",
audio: { url: "https://example.com/audio.mp3" },
},
10, // pollInterval seconds
300 // pollTimeout seconds
);

console.log("Transcription:", result.text);
`

Or submit first, then poll separately:

`typescript
// Start transcription
const stt = await client.speechToText.create({
model: "aai/slam-1",
audio: { url: "https://example.com/audio.mp3" },
});

// Poll for status
const status = await client.speechToText.getStatus(stt.generation_id);
console.log("Status:", status.status);
`

$3

`typescript
const audio = await client.textToSpeech.create({
model: "alibaba/qwen3-tts-flash",
text: "Hello, this is a test.",
voice: "Cherry",
});

console.log("Audio URL:", audio.audio.url);
`

$3

`typescript
const response = await client.images.generate({
model: "dall-e-3",
prompt: "A beautiful sunset over mountains",
size: "1024x1024",
});

console.log("Image URL:", response.data[0].url);
`

$3

Get balance and available models:

`typescript
// Get account balance
const balance = await client.account.getBalance();
console.log("Balance:", balance.balance);
console.log("Low balance:", balance.lowBalance);

// Get all available models
const models = await client.account.getModels();
console.log("Total models:", models.models?.length);

// Filter models by type
const chatModels = await client.account.getModelsByType("chat-completion");
console.log("Chat models:", chatModels.models?.length);

// Filter by provider
const openaiModels = await client.account.getModelsByDeveloper("openai");
console.log("OpenAI models:", openaiModels.models?.length);

// Find model by ID
const model = await client.account.getModelById("gpt-4o");
console.log("Model:", model);
`

$3

`typescript
const message = await client.messages.create({
model: "claude-sonnet-4.5",
messages: [{ role: "user", content: "Hello, Claude!" }],
max_tokens: 100,
});

console.log(message.content[0]?.text);
`

$3

`typescript
// Google Document AI OCR
const ocr = await client.ocr.google({
document: "https://example.com/document.png",
mimeType: "image/png",
});

console.log("Pages:", ocr.pages?.length);

// Mistral OCR
const mistralOcr = await client.ocr.mistral({
document: {
type: "image_url",
image_url: "https://example.com/image.jpg",
},
});

console.log("Text:", mistralOcr.pages?.[0]?.text);
`

$3

Process multiple requests asynchronously:

`typescript
const batch = await client.batch.create({
requests: [
{
custom_id: "request-1",
params: {
model: "claude-3-5-haiku-20241022",
messages: [{ role: "user", content: "Hello" }],
max_tokens: 100,
},
},
],
});

console.log("Batch ID:", batch.id);
console.log("Status:", batch.processing_status);

// Wait for completion
const result = await client.batch.waitForCompletion(batch.id, {
maxAttempts: 30,
pollInterval: 10,
});

// Parse results
const parsed = client.batch.parseResults(result as string);
`

$3

`typescript
// Search knowledge base
const knowledge = await client.bagoodex.searchKnowledge("Who is Nikola Tesla");
console.log("Title:", knowledge.title);

// Get weather
const weather = await client.bagoodex.getWeather("San Francisco, CA");
console.log("Temperature:", weather.temperature);

// Search images
const images = await client.bagoodex.searchImages("cute cats");
console.log("Results:", images.length);

// Search videos
const videos = await client.bagoodex.searchVideos("tutorial");
console.log("Results:", videos.length);
`

Available Models

| Provider | Models |
|----------|--------|
| OpenAI | gpt-4o, gpt-4o-mini, o1, o3-mini, gpt-5 |
| Anthropic | claude-sonnet-4.5, claude-opus-4.5 |
| Google | gemini-2.5-pro, gemini-2.0-flash |
| DeepSeek | deepseek-chat, deepseek-reasoner |
| Meta | llama-4-maverick, llama-4-scout |
| Video | openai/sora-2-t2v, alibaba/wan2.1-t2v-plus, luma/ray-2 |
| Music | elevenlabs/eleven_music, minimax/music-2.0, google/lyria2 |
| And more... | 400+ models total |

Environment Variables

| Variable | Description |
|----------|-------------|
| AIML_API_KEY | Your API key |
| AIML_API_BASE | Custom base URL (optional) |
| AIMLAPI_LOG` | Log level: debug, info, warn, error |

Documentation

- API Reference
- Examples
- Python SDK

License

Licensed under the Apache License 2.0.