Instructions to use GoldBean/goldbean-mcp with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use GoldBean/goldbean-mcp with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="GoldBean/goldbean-mcp")# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("GoldBean/goldbean-mcp", device_map="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use GoldBean/goldbean-mcp with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "GoldBean/goldbean-mcp" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "GoldBean/goldbean-mcp", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/GoldBean/goldbean-mcp
- SGLang
How to use GoldBean/goldbean-mcp with SGLang:
Install from pip and serve model
# Install SGLang from pip: pip install sglang # Start the SGLang server: python3 -m sglang.launch_server \ --model-path "GoldBean/goldbean-mcp" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "GoldBean/goldbean-mcp", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker images
docker run --gpus all \ --shm-size 32g \ -p 30000:30000 \ -v ~/.cache/huggingface:/root/.cache/huggingface \ --env "HF_TOKEN=<secret>" \ --ipc=host \ lmsysorg/sglang:latest \ python3 -m sglang.launch_server \ --model-path "GoldBean/goldbean-mcp" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "GoldBean/goldbean-mcp", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use GoldBean/goldbean-mcp with Docker Model Runner:
docker model run hf.co/GoldBean/goldbean-mcp
GoldBean MCP Server - x402 Micropaid AI API Marketplace
Access 51 Baidu AI tools via a single MCP endpoint. Pay-per-call with USDC micropayments. No Chinese phone number required.
What is GoldBean?
GoldBean is an MCP (Model Context Protocol) Server that provides pay-per-call access to 51 Baidu AI endpoints through a unified Streamable HTTP interface. Built on the x402 micropayment protocol, it enables developers to use Chinese AI services (OCR, TTS, ASR, Translation, ERNIE LLM, Vision, Image/Video Generation) without needing a Chinese phone number or Baidu Cloud account.
Key Features
- 51 AI Tools across 11 categories
- Streamable HTTP Transport (MCP spec 2024-11-05) - POST /mcp
- SSE Transport - GET /sse
- OpenAI-Compatible API - /v1/chat/completions, /v1/embeddings, /v1/images/generations
- x402 Micropayments - Pay per call with USDC on Base blockchain
- No KYC - Pay with crypto, PayPal, or Alipay
- Free Tier - 50 calls/day per IP + 100 credits on registration
Quick Start
MCP Client Configuration
Add to your MCP client config (Claude Desktop, Cursor, etc.):
{
"mcpServers": {
"goldbean": {
"url": "https://goldbean-api.xyz/mcp"
}
}
}
SSE Configuration
{
"mcpServers": {
"goldbean": {
"url": "https://goldbean-api.xyz/sse"
}
}
}
npm Package
npx goldbean-mcp
Tool Categories (51 tools)
| Category | Count | Examples |
|---|---|---|
| OCR | 10 | General, Handwriting, ID Card, Bank Card, License Plate, Table, Formula |
| NLP | 9 | Sentiment, Keyword Extraction, Text Correction, Entity Recognition, Summary |
| Vision | 10 | Object Detection, Face Detection, Logo Recognition, Animal, Plant, Ingredient |
| Image Processing | 4 | Background Removal, Image Enhancement, Cartoon Selfie, Style Transfer |
| Face and Body | 4 | Face Comparison, Body Analysis, Gesture Recognition |
| Voice | 2 | TTS (Text-to-Speech), ASR (Speech Recognition) |
| Translation | 1 | Multi-language Translation (200+ languages) |
| Video Generation | 2 | Text-to-Video, Image-to-Video |
| LLM | 3 | ERNIE-4.0, DeepSeek-R1, ERNIE-Speed |
| Bioinformatics | 1 | HelixFold Protein Structure Prediction |
| Free Tools | 5 | No payment required |
Pricing
| Plan | Price | Calls | Best For |
|---|---|---|---|
| Free | $0 | 50/day | Testing and evaluation |
| Per-call | $0.001-$0.08 | Pay as you go | Occasional use |
| Starter | $5 | 100 calls | Small projects |
| Monthly | $9.9 | Unlimited | Regular users |
| Quarterly | $25 | Unlimited | Power users |
| Yearly | $89 | Unlimited | Best value |
Payment Methods
- USDC on Base (x402 micropayments) - No KYC, instant settlement
- PayPal - Global coverage
- Alipay - China-friendly
OpenAI-Compatible API
import openai
client = openai.OpenAI(
base_url="https://goldbean-api.xyz/v1",
api_key="your-goldbean-api-key"
)
# Chat with ERNIE-4.0
response = client.chat.completions.create(
model="ernie-4.0",
messages=[{"role": "user", "content": "Hello!"}]
)
# Generate images
response = client.images.generations.create(
model="ernie-text-to-image",
prompt="A panda eating bamboo in a misty mountain"
)
Available on 9 Platforms
- Glama.ai - MCP Directory
- mcp.so - MCP Directory
- MCPize - MCP Hosting
- mcpservers.org - MCP Directory
- himcp.cn - Chinese MCP Directory
- MCP Registry - Official MCP Registry
- RapidAPI - API Marketplace
- Smithery - MCP Directory
- HuggingFace - AI Model Hub
Links
- Website: https://goldbean-api.xyz
- GitHub: wuzenghai616-lang/goldbean
- npm: goldbean-mcp
License
MIT License - see LICENSE