neurolink

Overview

NeuroLink serves as the vascular layer for AI intelligence, carrying streams of tokens, tool calls, memory, voice, and documents from LLM providers to connectors. It provides a production-ready solution with enterprise features like Redis memory, multi-provider failover, and automatic cost optimization. The platform includes 64+ built-in tools and MCP servers with tool routing, result caching, request batching, and multi-server management capabilities.

Try asking AI

After installing, here are 4 things you can ask your AI assistant:

• generate

  Generate AI content with various providers and options

  Try:Generate a summary of this documentTry:Create a blog post about AI trends
• stream

  Stream AI responses in real-time

  Try:Stream a response to this user queryTry:Process this input as a stream
• tts

  Text-to-speech conversion with multiple providers

  Try:Convert this text to speechTry:Generate audio from my written script
• stt

  Speech-to-text transcription with multiple providers

  Try:Transcribe this audio fileTry:Convert speech to text
• rag

  Retrieval-Augmented Generation with document processing

  Try:Find information in these documentsTry:Answer questions based on this research
• memory

  Manage conversation memory across sessions

  Try:Recall our previous conversationTry:Store this information for later
• autoresearch

  Run autonomous AI experiments

  Try:Optimize my model hyperparametersTry:Run experiments to improve performance
• process_files

  Process various file types (PDF, CSV, images)

  Try:Extract text from this PDFTry:Analyze this CSV data
• generate_ppt

  Create PowerPoint presentations

  Try:Generate a presentation about AI safetyTry:Create slides for my meeting
• generate_video

  Generate videos using AI models

  Try:Create a short promotional videoTry:Generate a video explanation of this concept
• generate_image

  Generate images using AI models

  Try:Create an illustration of a futuristic cityTry:Generate an image for this concept
• structured_output

  Generate structured JSON output

  Try:Extract data from this text in JSON formatTry:Generate a JSON response with specific schema

Note: Tool names inferred from documentation but exact signatures not provided. Many tools appear to be configuration options within the main generate() and stream() methods rather than standalone MCP tools.

Comparable tools

Installation

npm install @juspay/neurolink

{
  "mcpServers": {
    "neurolink": {
      "command": "node",
      "args": ["-e", "require('@juspay/neurolink')"],
      "env": {
        "NEUROLINK_API_KEY": "your-api-key"
      }
    }
  }
}

FAQ

How many AI providers does NeuroLink support?

NeuroLink supports 21+ AI providers including OpenAI, Anthropic, Google, AWS Bedrock, Azure, DeepSeek, NVIDIA NIM, and local models through LM Studio and llama.cpp.

What are the MCP capabilities in NeuroLink?

NeuroLink includes MCP servers with tool routing (6 strategies), result caching (LRU/FIFO/LFU), request batching, annotations, elicitation protocol, and multi-server management.

Compare neurolink with