Access multiple LLM models with one API key

Unified key management, quota control, and multi-model routing—focus on development.

Get API Key Developer Docs
API Usage Example
OpenAI-compatible format 
// Call LLM models with your API key
const response = await fetch('https://api.ithu.tw/v1/chat/completions', {
  method: 'POST',
  headers: {
    'Authorization': 'Bearer YOUR_API_KEY',
    'Content-Type': 'application/json'
  },
  body: JSON.stringify({
    'model': 'gpt-oss-120b',
    'messages': [
      { 'role': 'user', 'content': 'Hello, world!' }
    ]
  })
});
6+
Available Models
500+
Registered Users
99.9%
Uptime
<200ms
Avg. Latency

Key management designed for developers

Simplify API access so you can focus on product development.

Unified key management

Access all models with a single key—simplify deployment.

Secure key control

Quota management, usage tracking, and real-time monitoring.

Multi-model routing

Choose models based on needs—balance speed and cost.

OpenAI-compatible

Standard API format—easily integrate with existing tools.

Fast deployment

Get your key in 60 seconds and start developing immediately.

Usage analytics

Track API calls, token usage, and costs.

Supported integrations
Python SDK Node.js SDK REST API OpenAI SDK GitHub Actions Cloud Functions

Available models

Choose the model that fits your needs and start building.

bge-m3-embedding

BAAI BGE-M3 多語嵌入模型（dual-encoder），支援跨語言檢索與語義匹配；適用於向量資料庫、RAG 檢索、相似度搜尋與多語內容索引。
bge-m3-reranker

BAAI BGE Reranker v2 M3（cross-encoder）重排序模型，搭配嵌入檢索以提升最終排序品質；在長文本段落匹配與精準查詢場景表現突出。
coder

大型開發助理模型（coder），強化程式理解、重構、測試與工具使用；與 VS Code/Cline 等工作流整合良好，適合端到端的開發協作與自動化。 模型使用：nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16
gpt-oss-120b

開源 GPT 風格 120B 參數的大型語言模型，適合長上下文推理、程式輔助與通用生成。在內網自建環境可替代商用 API，支援多語與穩定的指令遵循，適合作為高品質預設模型。 我們現在已經換成 Mistral-Small-4-119B-2603 但為了往前相容，model名稱仍然使用 gpt-oss-120b
vibe

速度快，量大情況下約80t/s 使用模型：Nemotron-Cascade-2-30B-A3B
whisper-large-v3

Whisper Large v3（加速版）語音轉文字，強化多語音訊辨識與長音訊處理。
No matching models found. Try a different keyword.

Start in 60 seconds: get your API key and start developing

Simplify the process so you can focus on creating.

Get API Key