DEV Community: ailm32442型机器人

An AI Magazine You Can Never Finish Reading! Let Your AI Companion Take You Through the Changes of the World!

ailm32442型机器人 — Mon, 19 Jan 2026 19:17:47 +0000

Introduction

In the era of information explosion, traditional reading methods often consume too much time and effort. Therefore, I added a new extension — sap-rss — to the AI companion software — Super Agent Party. This extension helps us obtain updates through RSS subscriptions, automatically organizes them into a daily-updated AI magazine, and finally lets your AI companion read it aloud to you. Let’s take a look at how this is achieved!

Step One: Download Super Agent Party!

First, you need to download the free Super Agent Party! This is an open-source AI companion software. You can view its source code on GitHub.

If you are a Chinese user, you can download it from the ModelScope Community. Select the folder corresponding to the version number, then choose the version compatible with your operating system and click to download!

If you are an overseas user, you can download it from GitHub. Select the version suitable for your operating system and click to download!

If you are a tech-savvy geek, you can also choose the Docker version and check the usage instructions on GitHub.

Step Two: Install the RSS Extension

Open Super Agent Party, and you can see the installed extensions under the 【Developer】-> 【Extensions】subpage. Click the 【Install New Extension】button and select 【sap-rss】to install.

There are three ways to open the extension:

Open via sidebar: Click 【sap-rss】in the sidebar interface to open the extension.

After opening, you will see the RSS extension interface. You can click the + button to add new RSS subscriptions.

Open via independent window: Click the small independent window button at the top right of 【sap-rss】in the sidebar interface.

After opening, the RSS extension interface will appear as an independent window.

Open via browser window: Click the small browser window button at the top right of 【sap-rss】in the sidebar interface.

After opening, the RSS extension interface will appear as a browser window.

Step Three: Add RSS Subscriptions

In the RSS extension interface, you can click the + button to add new RSS subscriptions. You can enter the RSS URL and click the 【Add】button.

You can click the network button to the left of the + button to add some pre-organized RSS subscriptions we have prepared.

Step Four: Generate AI Translation, AI Summary, AI Podcast, AI Magazine

Click the buttons on the interface to generate AI translations, summaries, podcasts, and magazines based on your RSS subscriptions. Super Agent Party does not require you to pay for AI features; all AI computing power can be achieved through AI models you configure yourself. Therefore, you need to configure the corresponding LLM and TTS models in advance. How to use free computing power has already been introduced in the previous blog post — you can click here to view.

You can click the table of contents on the AI magazine to jump to the chapter you want to read.

Step Five: Let Your AI Companion Read It to You

Click the VRM pet button, then click the 【Play】button on the AI magazine to let your AI companion read it aloud to you.

Where to Find More RSS Subscriptions?

Although the RSS protocol is long-standing, it remains the most free and efficient way to access information on the internet today. Although many websites no longer prominently display the RSS icon, they are still available. To make your Super Agent Party-generated AI magazine content rich and diverse, here is a “Subscription Source Mining Guide” ranging from beginner to advanced, covering major global platforms.

1. Essential Tool: Browser Detection Plugin (RSSHub Radar)

This is the simplest and most recommended method for all users. You don’t need to know coding — just install a “radar” in your browser.

RSSHub Radar: This is an open-source browser extension (supports Chrome/Edge/Firefox). When you visit a website, if the site has RSS support or can generate RSS via RSSHub, its icon will light up and display a number.
- Applicable Scenarios: Almost all web pages.
- Acquisition Method: Search for RSSHub Radar on GitHub or directly download it from the Chrome Web Store.
- Global Applicability: Can detect not only domestic platforms like Bilibili and Weibo, but also international platforms like YouTube channels and Reddit subreddits.

2. Everything Can Be RSS: Embrace the Open-Source Project RSSHub

If you find that a website you like (e.g., Twitter/X, Instagram, Zhihu, Douyin, etc.) does not officially support RSS, what should you do? At this point, you need to bring out the “Swiss Army Knife” of the RSS world — RSSHub.

What is RSSHub: It is an open-source, community-driven project that can convert almost any web page content into an RSS subscription source through specific rules.
How to Use:
- Official/Public Instances: RSSHub provides a public domain name. You only need to construct the URL according to the documentation (e.g., https://rsshub.app/twitter/user/elonmusk).
- Self-Hosted Service (Recommended for Geeks): To ensure a more stable experience and access specific regional content, it is strongly recommended to deploy an RSSHub locally or on a server using Docker, just like deploying Super Agent Party.
- Docker Deployment: Just one command docker pull diygod/rsshub, combined with our sap-rss extension, the experience is simply amazing.

3. Native or Skill-Based Access to Major Global Platforms

Many international mainstream platforms actually have native RSS interfaces, though they are hidden deep within. Mastering these techniques allows you to easily subscribe to global news:

YouTube (Videos/Podcasts):
- No third-party tools needed. Open any channel page, view the source code to search for channel_id, then construct: https://www.youtube.com/feeds/videos.xml?channel_id=yourID.
- Combined with the AI companion’s summarization feature, you don’t even need to watch the video to know what the creator is talking about.
Reddit (Community Discussions):
- Very user-friendly — simply add .rss to the URL. For example: https://www.reddit.com/r/technology/.rss.
Medium (Technology/Writing):
- Add /feed to the end of the personal homepage URL. For example: medium.com/feed/@username.
Substack (Newsletters):
- In this era of newsletter popularity, almost all Substack pages only require adding /feed after the domain to obtain the full RSS feed.

4. Aggregation Platforms and Search Engines (Aggregators)

If you don’t know what to read, you can browse these specialized RSS “yellow pages” websites, which usually categorize by topic, making it easy for you to import in bulk:

Feedly / Inoreader: These are the two largest RSS reader service providers globally. Although they are primarily readers, their “search bar” is a powerful RSS search engine. Enter keywords (e.g., "AI News" or "SpaceX"), and they will list the most popular relevant RSS sources across the web.
Follow.is: A next-generation subscription browser with a very modern interface. Its Discover page features many high-quality user-created subscription lists (Lists) covering cutting-edge fields such as technology, design, and programming.
Feedspot: A database containing tens of thousands of RSS Feeds worldwide, categorized by industry and country — ideal for finding professional information in specific fields.

5. Awesome Lists on GitHub

For developers and tech-savvy users, GitHub itself is a massive resource repository. You can search for the following keywords to find community-maintained source lists (OPML files):

ALL-about-RSS
awesome-rss-feeds
chinese-independent-blogs (Chinese independent blog list)
engineering-blogs (Technical blogs from major global tech companies)

Tip: After obtaining the RSS link, don’t forget to test it in your browser first. If you see a screen full of code (XML format) or a download prompt, it means the link is valid — simply copy it into the sap-rss extension in Super Agent Party!

Conclusion

Through RSS subscriptions + AI companion, we can efficiently obtain updates, automatically organize them into a daily-updated AI magazine, and finally let your AI companion read it aloud to you. This way, we can access the latest news anytime and anywhere in our busy lives, while letting the AI companion help us organize and read these updates.If you have any questions, feel free to join:

QQ Group: 931057213
Discord: Discord link

Original blog link

© 2026 heshengtao, licensed under CC BY 4.0 International

Free to share, adapt, and use commercially, with attribution and a link to the original.

Challenge: Create an AI Companion for $0, Can I Succeed?

ailm32442型机器人 — Mon, 19 Jan 2026 19:15:38 +0000

Introduction

If you want an AI companion but don't want to spend a single cent, then this article is for you. Here, I will show you how to create an AI companion for free, allowing you to find companionship in the virtual world.

Step 1: Download the free Super Agent Party!

First, you need to download the free Super Agent Party! This is an open-source AI companion software. You can view its source code on GitHub.

If you are a user in China, you can download it from the ModelScope Community. Select the folder corresponding to the version number, then choose the version that matches your operating system, and click download!

If you are an overseas user, you can download it from GitHub. Select the version suitable for your operating system and click download!

If you are a tinkering geek, you can also choose the Docker version. Check the usage instructions on GitHub.

Step 2: Enable Your AI Companion to Think

You need a large language model, commonly referred to as an LLM. The LLM serves as the "brain" of your AI companion, determining its level of intelligence. There are many free LLM computing resources available on the market. Here, I recommend one that I frequently use: Ollama's cloud models. At least at the time I wrote this blog post, Ollama's cloud models are still free.

Download Ollama from the Official Website

Ollama official website: https://ollama.com/

In the top-right corner of the webpage, you'll see a "Download" button. Click it, select the version suitable for your operating system, and click download!

Free Ollama Cloud Models

You need to register an account on the Ollama website. After registering, log into your account within the Ollama application, and you'll be able to use Ollama's cloud models.
On the "Models" page of the Ollama website, click the cloud tab. The models under this tab are free and do not consume your local computing resources.

You can choose a model you like, click on it, and copy the CLI command, for example, ollama run deepseek-v3.1:671b-cloud. Paste it into your terminal and press Enter to start using Ollama's cloud model. Important! Make sure the model has the cloud suffix.

Using Ollama Cloud Models in Super Agent Party

Open Super Agent Party, go to the Models page, and under the Model Providers subpage, select Add Provider.

In the modal window, select ollama, then click the Confirm Add button.

Click the magnifying glass button on the card to retrieve a list of all models within your Ollama. If the retrieval fails, you might have forgotten to open the Ollama application.

From the model list, select the model you just copied, for example, deepseek-v3.1:671b-cloud.

Now, return to the chat interface, and you can happily chat with your AI for free!

Step 3: Enable Your AI Companion to Hear

You need a speech recognition model, commonly referred to as ASR. ASR serves as the "ears" of your AI companion, determining whether it can understand what you say.

Here are two free methods:

Free Local Small Model

Super Agent Party comes with a built-in free speech recognition model. Simply go to the Models page, select Speech Recognition Model, and then choose Sherpa ONNX to use the built-in speech recognition model. This model does not require GPU resources; it only uses CPU and memory, so it can run on an ordinary laptop without the need to invest in expensive hardware.

Open Super Agent Party, click on the Models page, then in the Speech Recognition Model subpage, select Sherpa ONNX.
Click the download button in the Sherpa ONNX Sense Voice card below. Users in China can choose to download from ModelScope, while overseas users can download from Hugging Face.
Once the download is complete, you can enable speech recognition and start chatting with your AI using voice.

Free Browser Built-in Speech Recognition API

You can use the free speech recognition API built into browsers, known as the Web Speech API. You need to open a browser to serve as the "ears" of your AI companion and use the Web Speech API for speech recognition. This method is free but requires an internet connection.

Open Super Agent Party, click on the Models page, then in the Speech Recognition Model subpage, select Web Speech API.
Enable speech recognition. Super Agent Party will automatically open a browser, and now you can use the Web Speech API in the browser for speech recognition.

Step 4: Enable Your AI Companion to Speak

You need a speech synthesis model, commonly referred to as TTS. TTS serves as the "mouth" of your AI companion, determining whether it can communicate with you using voice.

Here are two free methods:

Free System TTS

Both Windows and macOS systems come with built-in speech synthesis models. The voice quality is average, but the speed is very fast, making it suitable for quick testing. For macOS, choose voices labeled with 【Siri/Premium】, such as 【Siri/Premium】Meijia, which is the familiar Siri voice!

Open Super Agent Party, click on the Models page, then in the Speech Synthesis Model subpage, select System TTS.
Choose your system TTS, such as 【Siri/Premium】Meijia.
Enable speech synthesis.

Free EdgeTTS

EdgeTTS is a free TTS interface provided by Microsoft, offering better voice quality but requiring an internet connection. It is important to note that Microsoft permits small-scale or personal classroom use, but commercial and profit-driven use is not allowed. You can find relevant explanations on Microsoft's official website. For commercial use, please use Azure TTS.

Open Super Agent Party, click on the Models page, then in the Speech Synthesis Model subpage, select EdgeTTS.
Choose a voice for the corresponding language, such as xiaoxiao.
Enable speech synthesis.

Step 5: Enable Your AI Companion to Be Seen

To make your AI companion appear as a 3D character on your desktop, you need to enable the VRM Desktop Pet Robot.

Free VRM Models

If you want more free 3D models, you can download them from vroidhub or 模之屋.

Open Super Agent Party, click on the Robots page, then in the VRM Desktop Pet Robot subpage, click to start!
Use the left mouse button to rotate, the scroll wheel to zoom, the right mouse button to pan, and the first button in the upper right corner to drag the entire window.
If you want to record a digital human video with a transparent background or capture the desktop pet during a live stream, add http://127.0.0.1:3456/vrm.html as a video source in the browser source of your recording software.

If you also want your AI companion to appear as a 2D character, you will need to install the Live2D extension. I will cover this in detail in the extensions section. Stay tuned and bookmark this blog series.

Conclusion

Finally! We've completed an AI companion without spending a single penny! If you have any questions, feel free to join:

QQ Group: 931057213
Discord: Discord link

Original blog link

© 2026 heshengtao, licensed under CC BY 4.0 International

Free to share, adapt, and use commercially, with attribution and a link to the original.