The BEST 10 n8n Apps Released in 2025 (I Wish I Knew Sooner)

Key Concepts

Google Gemini, 11 Labs, Ampify (Web Scraping), AirTop (Browser Automation), Perplexity (LLM with Web Search), Reranker (with Vector Databases/RAG), Mistral AI (PDF to Text), Think Tool (Anthropic), DeepSeek (LLM), Open Router (LLM Aggregator), MCP (Multi-Client Protocol).

Google Gemini Integration

• Main Topic: Showcasing Google Gemini integration within NADN as a game-changer.
• Key Points:
  • Requires updating NADN instance to access.
  • Offers functionalities like audio analysis, transcription, document analysis, video generation, and image generation.
  • Demonstrates video generation (V2) and image generation capabilities with examples.
• Examples:
  • Generating a sci-fi themed image.
  • Generating a video.

11 Labs Integration

• Main Topic: Highlighting the 11 Labs community edition integration for text-to-speech and speech-to-text conversion.
• Key Points:
  • Community editions are created by the app developer and are free to install.
  • Focuses on converting text to speech.
  • Voice cloning is possible.
• Examples:
  • Converting text to speech using a deep voice model.

Ampify (Web Scraping) Integration

• Main Topic: Explaining the Ampify integration for web scraping and its potential applications.
• Key Points:
  • Web scraping allows extracting data from websites.
  • Use cases include extracting contact information from Google Maps for cold emailing and finding viral ideas on social media.
  • Provides $1,500 of credits per month, equating to approximately 1,000 free web scrapes.
  • Includes a "wait for finish" function that optimizes waiting time during web scraping.
• Examples:
  • Web scraping the presenter's agency website to extract text and social media handles.
• Technical Terms:
  • Web Scraping: Extracting data from websites.

AirTop (Browser Automation) Integration

• Main Topic: Demonstrating AirTop integration for browser automation through AI agents.
• Key Points:
  • Allows running browser automation directly through an AI agent or NADN.
  • Actions include loading URLs, querying, typing, and ending sessions.
• Examples:
  • Using an AI agent to find Nike shoes on eBay.

Perplexity (LLM with Web Search) Integration

• Main Topic: Introducing Perplexity as a large language model with web searching capabilities.
• Key Points:
  • Similar to ChatGPT or Google Gemini but excels in web searching.
  • Provides citations and URLs for the information it retrieves.
• Examples:
  • Searching for the latest funding rounds in the AI automation space.

Reranker (with Vector Databases/RAG) Integration

• Main Topic: Explaining the Reranker integration for improving the accuracy of results from vector databases in a Retrieval Augmented Generation (RAG) system.
• Key Points:
  • RAG system: A database trained on the history of a company.
  • Reranker filters retrieved results to provide the most accurate answers.
• Examples:
  • Using a RAG database for a fictitious AI automation company's FAQ section.
  • Demonstrating how Reranker improves the quality of answers to pricing inquiries.
• Technical Terms:
  • RAG (Retrieval Augmented Generation): A system that retrieves information from a database to augment the generation of responses.
  • Vector Database: A database that stores data as vectors, enabling similarity searches.

Mistral AI (PDF to Text) Integration

• Main Topic: Showcasing the Mistral AI integration for extracting text from PDF documents.
• Key Points:
  • Converts PDF files to text for further processing.
  • Enables updating CRM or QuickBooks with data from PDFs.
• Examples:
  • Extracting text from a service agreement PDF.

Think Tool (Anthropic) Integration

• Main Topic: Explaining the Think Tool integration with Anthropic's Claude for improved AI agent reasoning.
• Key Points:
  • Two types of thinking: extended thinking (before generating a response) and the think tool (during response generation).
  • The think tool adds a step for the AI agent to pause and consider if it has all the necessary information.
• Examples:
  • Demonstrating how the think tool helps an AI agent reschedule a conflicting calendar event by prompting for more information.

DeepSeek and Open Router (LLMs) Integrations

• Main Topic: Briefly introducing DeepSeek as a large language model and Open Router as an aggregator for multiple LLMs.
• Key Points:
  • DeepSeek is a cheaper alternative to other LLMs.
  • Open Router provides access to various LLMs through a single subscription.

MCP (Multi-Client Protocol) Integration

• Main Topic: Discussing the MCP integration for two-way communication between AI agents, but noting its limited functionality in NADN.
• Key Points:
  • MCP allows communication between NADN and external AI agents like Claude.
  • The NADN implementation lacks full MCP functionality, such as merging multiple nodes into one.
  • Sending data from NADN to an external AI agent and back is less efficient than using NADN sub-workflows.
• Examples:
  • Scheduling a coffee event on a calendar using Claude via MCP.

Conclusion

The video provides a detailed overview of ten new and noteworthy app integrations within NADN. These integrations span various functionalities, including AI-powered content generation (Google Gemini), voice synthesis (11 Labs), web scraping (Ampify), browser automation (AirTop), enhanced search capabilities (Perplexity, Reranker), document processing (Mistral AI), improved AI reasoning (Think Tool), and multi-client protocol communication (MCP). While some integrations offer complete functionality, others, like MCP, are still under development. The presenter emphasizes the potential of these integrations to streamline workflows, automate tasks, and enhance the capabilities of AI agents within the NADN ecosystem.