GoldBean API
x402/goldbean-apix402 service listed on awesome-x402: GoldBean API [x402: $0.001/call on eip155:8453, category=utility]
Tools · 51
Check GoldBean API service status, version, and uptime. Returns: { status, version, uptime, endpoints, free_calls_remaining }.
General-purpose OCR — extract printed text from images. Supports Chinese, English, Japanese, and 20+ languages with >90% accuracy. Returns: { words_result: [{ words: "text" }], words_result_num, langu…
High-accuracy OCR for low-quality, blurry, or complex documents. Returns structured text with confidence scores. Use when baidu_ocr results are insufficient.
Table extraction from images — returns structured rows/columns as JSON array. Handles merged cells and multi-page tables.
Chinese ID card (身份证) recognition. Extracts: name, gender, nationality, birth date, address, ID number. Supports front/back side detection.
Handwritten text recognition — extract handwritten Chinese or English text from images of notes, forms, or letters.
QR code and barcode recognition — decode 1D/2D barcodes from images. Supports QR, Data Matrix, PDF417, Code128, and more.
Bank card recognition — extract card number, bank name, and card type from photos of bank cards.
Chinese business license (营业执照) recognition. Extracts: company name, unified social credit code, legal representative, registered capital, establishment date, business scope.
OCR optimized for web images and screenshots — handles compressed, low-resolution images from web pages. Returns text with positional data.
Advanced OCR with DeepSeek-OCR model — handles complex layouts, multi-column text, mixed text-and-image content, and mathematical formulas. Supports custom extraction prompts.
Document parsing with PaddleOCR-VL — performs layout analysis, reading order detection, and structure extraction. Returns structured document with headings, paragraphs, tables, and figures identified.
General-purpose OCR with Qianfan-OCR model — supports up to 32k context length, ideal for long documents and multi-page scans.
Text-to-Speech synthesis — convert text to natural-sounding audio. Supports Chinese (Mandarin/Cantonese) and English. Returns MP3 audio binary. Voice options: 0=female, 1=male, 3=emotional, 4=emotiona…
Automatic Speech Recognition — transcribe audio to text. Supports Chinese (Mandarin), English, and Cantonese. Returns: { result: ["transcribed text"] }.
Chat with ERNIE LLM — Baidu's flagship large language model (comparable to GPT-4). Excellent at Chinese language tasks, reasoning, code generation, and multi-turn conversation. OpenAI-compatible. Retu…
Deep reasoning with DeepSeek-R1 — chain-of-thought reasoning for math, logic, and complex analysis. Returns both reasoning trace and final answer. Returns: { reasoning: "step-by-step", result: "answer…
Vision-language model — describe and analyze images using ERNIE-4.5-VL or Qwen3-VL. Supports visual question answering, image description, and OCR-like tasks. Returns: { result: "description/answer" }…
Multi-language translation powered by Baidu Translate. Supports 200+ languages including Chinese, English, Japanese, Korean, French, German, Spanish, Arabic, etc. Returns: { trans_result: [{ src, dst …
General image recognition — identify objects, scenes, concepts, and activities in images. Returns categorized labels with confidence scores.
Object detection — locate and identify multiple objects in an image with bounding boxes. Returns: { results: [{ name, score, location: {left, top, width, height} }] }.
Landmark recognition — identify famous landmarks, buildings, and monuments from photos. Returns: { result: { landmark, score } }.
Plant species recognition — identify plants, flowers, and trees from photos. Returns: { result: [{ name, score, baike }], optional Wikipedia info }.
Animal species recognition — identify animals, birds, and insects from photos. Returns: { result: [{ name, score, baike }] }.
Food/cuisine recognition — identify dishes and food items from photos. Returns: { result: [{ name, score, calories }] }.
Brand logo recognition — identify brand logos in images. Returns: { result: [{ name, score, location }] }.
Car model recognition — identify car make, model, and year from photos. Returns: { result: [{ name, score, year }] }.
Fruit and vegetable recognition — identify fresh produce from photos. Returns: { result: [{ name, score }] }.
Vehicle detection and counting — detect all vehicles in an image (cars, trucks, buses, motorcycles) with bounding boxes. Useful for parking lot management and traffic analysis.
Face detection and attribute analysis — returns face location, age estimate, gender, emotion, beauty score, glasses, face shape, and expression for each detected face.
1:1 face verification — compare two faces to determine if they belong to the same person. Returns similarity score (0-100) and match threshold. Returns: { score, threshold, is_same_person }.
Human body detection and analysis — detect people, recognize posture/pose, and count individuals in images. Returns: { person_num, person_info: [{ location, attributes }] }.
Hand gesture recognition — identify gestures from photos: fist, open palm, peace sign, thumbs up, OK sign, etc. Returns: { result: [{ classname, probability, location }] }.
AI-powered image quality enhancement — denoise, deblur, dehaze, colorize black-and-white photos, and enhance contrast. Returns enhanced image as base64.
AI image generation from text prompts using Qwen-Image model. Excellent at rendering Chinese text within images. Supports various art styles. Returns: { image_url, seed }.
AI image editing — modify existing images with text instructions using Qwen-Image-Edit model. Can add/remove objects, change style, or transform content. Returns: { image_url }.
AI video generation from image + text prompt using MuseSteamer model. Submit a reference image and motion description to generate a short video clip. Async: returns task_id for polling. Returns: { tas…
Query video generation task status. Use the task_id returned by baidu_video_gen. Returns: { task_status, video_url } when complete, or { task_status: "running" } while processing.
Chinese NLP lexical analysis — word segmentation (分词) and part-of-speech tagging. Returns: { items: [{ word, pos_tag, ne }], basic_words }.
Sentiment analysis — classify text as positive (2), negative (0), or neutral (1). Returns: { items: [{ sentiment, confidence, prop }], text }.
Automatic text summarization — condense long articles into concise summaries. Returns: { summary: "condensed text" }.
Chinese text error correction — detect and fix typos, grammatical errors, and character mistakes in Chinese text. Returns: { corrected_text, errors: [{ original, corrected }] }.
Keyword extraction — identify the most important keywords from an article. Returns: { keywords: [{ word, score }] }.
Word embedding — convert words or short phrases to dense vector representations (768-dim) for similarity computation, clustering, and ML tasks.
Text embedding for semantic search, clustering, and RAG — convert text to dense vectors. Supports embedding-v1, BGE, and Qwen3 models. Use || to separate multiple texts for batch embedding.
Document reranking — rerank a list of documents by relevance to a query. Essential for RAG pipelines to improve retrieval accuracy. Returns: { results: [{ index, relevance_score }] }.
Text content moderation — detect spam, porn, violence, politically sensitive content, and abuse in text. Returns: { result: [{ type, probability, hit }] }.
Image content moderation — detect inappropriate or unsafe content in images (porn, violence, political sensitivity). Returns: { result: [{ type, probability }] }.
Protein 3D structure prediction using HelixFold3 (AlphaFold3-class model). Predict protein structure from amino acid sequence for drug discovery and bioinformatics research. Returns: { pdb_url, confid…
Register for a free GoldBean API key. Returns your GB_ API key with 100 free credits. No credit card needed. Returns: { api_key, credits, message }.
Check your remaining API credits, usage statistics, and subscription status. Returns: { credits_remaining, calls_today, plan, free_calls_remaining }.
Similar MCP servers embedding-nearest
How to use
Add to your Claude Desktop / Cursor / Cline MCP config:
{
"mcpServers": {
"goldbean_api": {
"url": "https://goldbean-api.xyz/mcp",
"transport": "http"
}
}
}