Introducing Yunxin

Today we're announcing Yunxin — a unified AI API gateway that gives you OpenAI-compatible access to models from OpenAI, Anthropic, Google, DeepSeek, and 20+ other providers.

The Problem

Building with multiple AI providers means dealing with:

• Different SDKs — Each provider has their own client library
• Different APIs — Message formats, streaming implementations, and error handling vary
• Multiple dashboards — Tracking usage across providers is fragmented
• Vendor lock-in — Switching models requires code changes

Our Solution

Yunxin provides a single OpenAI-compatible API endpoint. Use your existing OpenAI SDK — just change the base URL:

from openai import OpenAI

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

# Use any supported model
response = client.chat.completions.create(
    model="model-id",  # Any model from any provider
    messages=[{"role": "user", "content": "Hello!"}]
)

No new SDKs. No code refactoring when switching models.

Supported Providers

Yunxin integrates directly with 24+ AI providers:

Major Providers:

• OpenAI — Chat, vision, embeddings, image generation, audio
• Anthropic — Claude with Extended Thinking, Computer Use
• Google — Gemini with long context and multimodal
• DeepSeek — Reasoning models with visible thinking
• xAI — Grok models
• Mistral — Open-source and commercial models

Chinese Providers:

• Alibaba (Tongyi) — Qwen models
• Zhipu AI — GLM models
• Moonshot — Kimi models
• MiniMax — Multimodal generation
• ModelScope — ModelScope models

Local & Open Source:

• Ollama — Local model serving
• LM Studio — Desktop inference
• vLLM — High-throughput serving
• Hugging Face — Serverless inference

And more: Azure, Cloudflare, Groq, NVIDIA NIM, OpenRouter, Volcengine, Z.AI

View all providers →

Key Features

OpenAI Compatibility

Yunxin implements the full OpenAI API surface:

• Chat Completions API
• Streaming (SSE)
• Function Calling
• Vision (image understanding)
• Embeddings
• Batch API
• Realtime API
• Responses API

Reasoning Models

First-class support for thinking/reasoning models with visible reasoning content. Access the thinking process via reasoning_content in responses.

Learn about reasoning →

Provider Architecture

Each provider has an independent adapter:

• 100% API fidelity — All provider-specific features work
• Optimal format — We use each provider's best API format
• Independent updates — New features added without affecting others

Unified Management

• One dashboard — All usage across all providers
• One bill — Simplified billing and invoicing
• Budget controls — Set limits and alerts
• Analytics — Track usage by model, API key, or project

Getting Started

1. Sign Up

Create an account at api.yuhuanstudio.com. New accounts receive starter credits.

2. Get an API Key

Navigate to Dashboard → API Keys → Create API Key

3. Make Your First Request

curl https://api.yuhuanstudio.com/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "model-id",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

Read the quickstart guide →

Available Models

Models and pricing are synced automatically from each provider. Use the Models API to query available models:

# List all models
curl https://api.yuhuanstudio.com/v1/models \
  -H "Authorization: Bearer YOUR_API_KEY"

# Filter by provider
curl https://api.yuhuanstudio.com/v1/models?provider=anthropic \
  -H "Authorization: Bearer YOUR_API_KEY"

# Filter by capability
curl https://api.yuhuanstudio.com/v1/models?capability=thinking \
  -H "Authorization: Bearer YOUR_API_KEY"

Check the Models dashboard for current availability and pricing.

Learn about the Models API →

Pricing

Yunxin uses pay-as-you-go pricing based on actual usage:

• Pricing is transparent — based on upstream costs plus a service margin
• View pricing for each model in the Models dashboard
• Set budget limits and alerts in Settings → Billing

Learn about pricing →

What's Next

We're actively developing:

• Smart routing — Automatically select optimal models
• Response caching — Reduce costs for repeated requests
• Enhanced analytics — Deeper cost and performance insights

Start Building

Yunxin is available now. Sign up and start building with any AI model through a single API.

Get Started →

Questions? Check out our documentation or reach out at [email protected].