惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

量子位
博客园_首页
罗磊的独立博客
云风的 BLOG
云风的 BLOG
J
Java Code Geeks
Last Week in AI
Last Week in AI
D
DataBreaches.Net
Jina AI
Jina AI
博客园 - Franky
大猫的无限游戏
大猫的无限游戏
Apple Machine Learning Research
Apple Machine Learning Research
V
V2EX
D
Docker
MongoDB | Blog
MongoDB | Blog
B
Blog RSS Feed
让小产品的独立变现更简单 - ezindie.com
让小产品的独立变现更简单 - ezindie.com
宝玉的分享
宝玉的分享
Engineering at Meta
Engineering at Meta
The Cloudflare Blog
博客园 - 三生石上(FineUI控件)
有赞技术团队
有赞技术团队
人人都是产品经理
人人都是产品经理
H
Help Net Security
T
The Blog of Author Tim Ferriss

DEV Community

Authentication Security Deep Dive: From Brute Force to Salted Hashing (With Java Examples) Why AI Systems Don’t Fail — They Drift Spilling beans for how i learn for exam😁"Reinforcement Learning Cheat Sheet" I Replaced Chrome with Safari for AI Browser Automation. Here's What Broke (and What Finally Worked) How Python Borrows Other People's Work The $40 Architecture: Processing 1 Billion API Requests with 99.99% Uptime Vibe Coding: A Workflow Guide (From Zero to SaaS) Most webhook security guides protect the wrong side. The scary part is delivery. Headless CMS for TanStack Start: Build a Blog with Cosmic EU Age Verification App "Hacked in 2 Minutes" — What Actually Happened Comfy Cloud’s delete function does not actually remove files Running AI Models on GPU Cloud Servers: A Beginner Guide Event-driven media intelligence with AWS Step Functions and Bedrock I scored 500 AI prompts across 8 quality dimensions — here's what broke How to Call Google Gemini API from Next.js (Free Tier, No Backend Needed) The Portal Protocol: Reclaiming Human Connection in the Age of AI How to Fix Your Team's Scattered Knowledge Problem With a Self-Hosted Forum Intro to tc Cloud Functors: A Graph-First Mental Model for the Modern Cloud Designing Multi-Tenant Backends With Both Ownership and Team Access I Built a Neumorphic CSS Library with 77+ Components — Here's What I Learned PostgreSQL Performance Optimization: Why Connection Pooling Is Critical at Scale Cómo construí un SaaS multi-rubro para gestionar expensas en Argentina con FastAPI + Vue 3 🚀 I Built an Ethical Hacking Scanner Tool – Open Source Project I Replaced /usage and /context in Claude Code With a Single Statusline A Pythonic Way to Handle Emails (IMAP/SMTP) with Auto-Discovery and AI-Ready Design I Collected 8.9 Million Polymarket Price Points — Here's What I Found About How Markets Really Move EcoTrack AI — Carbon Footprint Tracker & Dashboard Everyone's Using AI. No One Agrees How. 5 self-hosted ebook managers worth trying in 2026 Building Your First AI Agent with LangChain: From Chatbot to Autonomous Assistant
Tencent just released a RAG framework and nobody's talkin...
retrovirusre · 2026-05-21 · via DEV Community

retrovirusretro

In April 2026, Tencent's WeChat team released WeKnora as open source. MIT licensed. Ollama support built-in. Almost zero English content about it.

I spent a few days setting it up, writing the first English integration guide, and comparing it to the alternatives. Here's what I found.


What is WeKnora?

WeKnora is the RAG framework that powers WeChat's Dialog Open Platform — production-tested at a scale most of us will never reach.

At its core it does what every RAG tool does: upload documents, ask questions, get answers grounded in your content.

But it adds two things I haven't seen elsewhere:

1. Autonomous reasoning agent

When you ask a complex question, WeKnora doesn't just search. It plans.

"Compare the pricing strategy in document A with the market analysis in document B" gets decomposed into sub-queries before any retrieval happens. Most RAG tools dump a random mix of chunks into the LLM and hope for the best. WeKnora's agent actually thinks about how to answer before searching.

2. Self-updating knowledge base

Point WeKnora at a URL or folder, set a refresh interval, and it monitors the source and updates the knowledge base automatically when content changes. For internal docs, product catalogs, or anything that evolves — this is genuinely useful.


Setup in 5 Minutes

Two modes. Pick based on what you already have running.

If you already have Ollama running:

"bash
git clone https://github.com/retrovirusretro/weknora-english-guide
cd weknora-english-guide
cp .env.example .env
docker compose up -d"

WeKnora joins your existing Docker network. No duplicate Ollama container.

Fresh install (includes Ollama):
"docker compose -f docker-compose.standalone.yml up -d
docker exec -it weknora-ollama ollama pull llama3"

Open http://localhost:8083 → admin / weknora123 → connect Ollama → upload a PDF → ask a question.

*FastGPT prohibits commercial SaaS resale.

When to pick WeKnora over RAGFlow:

You need the reasoning agent for complex multi-document questions
MIT license matters (embedding in a commercial product)
You want the self-updating KB feature
When to pick RAGFlow instead:

Your PDFs have complex layouts (tables, multi-column, images)
You want a larger English community with more answered questions

n8n Integration
WeKnora exposes a REST API. Connect it to n8n for automation pipelines:
Webhook → WeKnora /api/query → Slack / Email / Notion

A ready-to-import n8n workflow JSON is in the repo:
examples/with-n8n/weknora-query-workflow.json

Import it in n8n → Workflows → Import from file. One click, working webhook.

Why Almost No English Content?
The WeKnora community is on WeChat groups and Zhihu. The maintainers write English READMEs but the tutorial ecosystem never crossed over.

Same story with FastGPT (27K stars), MaxKB (20K stars), DB-GPT (17K stars). Massive Chinese communities, almost nothing in English.

I'm documenting all of them:
→ chinese-ai-tools-english-guide

Full Guide
Everything in this post plus Ollama model selection, production deployment with Nginx + SSL, and the WeKnora vs RAGFlow deep-dive:

→ github.com/retrovirusretro/weknora-english-guide

Have you tried WeKnora? Curious if others run into setup issues I haven't documented yet.