About NexusBridge
NexusBridge is an MCP server published by guhank in the RAG category: access 200+ AI models and 10 service categories through a single NexusBridge API key. Run LLM chat completions, code generation, image generation, text embeddings, web search, document parsing, sentiment analysis, translation, summarization, and sandboxed code execution — all billed per call with transparent pricing. Credits never expire. No subscriptions required. Get your API key at https://syntss.com.
Six months or more without a commit doesn't mean the server is broken, but check the open issues before depending on it in production.
Install
git clone https://github.com/guhank/nexusbridge-mcp-serverThis server has no ConduID identity, so agent calls to it are not receipted. Pin the version you install and review the source before granting it credentials.
Tools (3)
nexusbridge_balanceCheck your current credit balance (USD) and the number of active services available. Use before expensive operations to verify sufficient funds.
Parameters not published.
nexusbridge_catalogRetrieve the full service catalog with slugs, descriptions, per-call prices, and your current credit balance. Call this first to discover available services.
Parameters not published.
nexusbridge_executeExecute any NexusBridge service by its slug. Pass service-specific parameters. Credits are deducted automatically. Returns the upstream provider's response.
Parameters not published.
Ask AI
Ask AI about NexusBridge
Powered by Claude · Grounded in docs
Security checks
- ✓README presentThe repository ships documentation.
5 checks have not been run yet.
README
NexusBridge MCP Server
A Model Context Protocol server that gives AI agents access to 200+ AI models and 10 service categories through a unified, metered API. One API key, one integration, transparent per-call pricing.
API Base: https://syntss.com
Smithery: smithery.ai/servers/guhank/nexusbridge-mcp-server
GitHub: github.com/guhank/nexusbridge-mcp-server
Tools
| Tool | Description |
|---|---|
nexusbridge_catalog |
Retrieve the full service catalog with slugs, descriptions, per-call prices, and your current credit balance. Call this first to discover available services. |
nexusbridge_execute |
Execute any NexusBridge service by its slug. Pass service-specific parameters. Credits are deducted automatically. Returns the upstream provider's response. |
nexusbridge_balance |
Check your current credit balance (USD) and the number of active services available. Use before expensive operations to verify sufficient funds. |
nexusbridge_catalog
Returns every available service with its slug, name, category, description, and per-call price in USD. Also returns your current credit balance.
Parameters: None
Example response:
{
"services": [
{ "slug": "llm-chat", "name": "LLM Chat", "category": "ai", "pricePerCall": 0.005 },
{ "slug": "web-search", "name": "Web Search", "category": "search", "pricePerCall": 0.01 }
],
"creditBalance": 24.50
}
nexusbridge_execute
Execute a service by passing its slug and service-specific parameters.
Parameters:
| Parameter | Type | Required | Description |
|---|---|---|---|
slug |
string | Yes | Service slug from the catalog (e.g., llm-chat, web-search, sentiment) |
params |
object | No | Service-specific parameters (see examples below) |
Examples:
// LLM Chat — use any of 200+ models
{ "slug": "llm-chat", "params": { "messages": [{"role": "user", "content": "Explain quantum computing"}], "model": "deepseek/deepseek-chat-v3-0324" } }
// Sentiment Analysis
{ "slug": "sentiment", "params": { "text": "I love this product! Best purchase I've ever made." } }
// Web Search
{ "slug": "web-search", "params": { "query": "latest AI news March 2026" } }
// Text Summarization
{ "slug": "summarize", "params": { "text": "Long article text here..." } }
// Translation
{ "slug": "translation", "params": { "text": "Hello world", "to": "es" } }
// Code Generation
{ "slug": "llm-code", "params": { "prompt": "Write a Python function to calculate fibonacci numbers" } }
// Image Generation
{ "slug": "image-gen", "params": { "prompt": "A serene mountain landscape at sunset" } }
// Text Embedding
{ "slug": "text-embedding", "params": { "text": "semantic search query" } }
Example response:
{
"success": true,
"service": "llm-chat",
"charged": 0.005,
"remainingBalance": 24.495,
"result": {
"choices": [{ "message": { "content": "Quantum computing uses..." } }]
}
}
nexusbridge_balance
Returns your credit balance and the count of active services.
Parameters: None
Example response:
{
"creditBalance": 24.50,
"serviceCount": 10
}
Available Services
| Service | Slug | Price/Call | Description |
|---|---|---|---|
| LLM Chat | llm-chat |
$0.005 | Chat completions with GPT-4o, Claude, Gemini, DeepSeek, Llama, and 200+ models via OpenRouter |
| Code Generation | llm-code |
$0.008 | Code generation, refactoring, and completion optimized for development tasks |
| Image Generation | image-gen |
$0.05 | Create images from text prompts using DALL-E, SDXL, and Flux models |
| Text Embedding | text-embedding |
$0.001 | Vector embeddings for RAG pipelines, semantic search, and clustering |
| Web Search | web-search |
$0.01 | Structured web search results with titles, URLs, and snippets |
| Document Parser | doc-parser |
$0.02 | Extract text and structured data from PDFs, images, and documents |
| Sentiment Analysis | sentiment |
$0.003 | Classify text sentiment (positive/negative/neutral) with confidence scores and emotion detection |
| Translation | translation |
$0.005 | Neural machine translation across 100+ languages |
| Summarization | summarize |
$0.008 | Configurable-length text summaries from long documents |
| Code Execution | code-exec |
$0.015 | Run code in sandboxed Python, Node.js, and Go environments |
Installation
Smithery (Recommended)
Install directly from Smithery with one click. Configure your NexusBridge API key when prompted.
Claude Desktop
Add to your claude_desktop_config.json:
{
"mcpServers": {
"nexusbridge": {
"command": "npx",
"args": ["-y", "nexusbridge-mcp-server", "--stdio"],
"env": {
"NEXUSBRIDGE_API_KEY": "nb_sk_your_key_here"
}
}
}
}
Cursor / Windsurf / Other MCP Clients
Use the same configuration pattern — set NEXUSBRIDGE_API_KEY as an environment variable and run via npx -y nexusbridge-mcp-server --stdio.
Docker
docker build -t nexusbridge-mcp .
docker run -p 8000:8000 \
-e NEXUSBRIDGE_API_KEY=nb_sk_your_key \
nexusbridge-mcp
Manual
git clone https://github.com/guhank/nexusbridge-mcp-server.git
cd nexusbridge-mcp-server
npm install
NEXUSBRIDGE_API_KEY=nb_sk_your_key npm start
Usage
-
Get an API key — Purchase credits at syntss.com. Credits start at $10. You'll receive an
nb_sk_...API key. -
Install the MCP server — Use Smithery, Claude Desktop config, or run manually (see Installation above).
-
Discover services — Your AI agent calls
nexusbridge_catalogto see all available services, pricing, and your credit balance. -
Execute services — Call
nexusbridge_executewith the service slug and parameters. Credits are deducted per call. -
Monitor spending — Call
nexusbridge_balanceto check remaining credits at any time.
Typical Agent Workflow
Agent → nexusbridge_catalog → "I see 10 services, balance is $24.50"
Agent → nexusbridge_execute(llm-chat) → "Got response from DeepSeek, charged $0.005"
Agent → nexusbridge_execute(sentiment)→ "Text is positive (0.94 confidence), charged $0.003"
Agent → nexusbridge_balance → "Balance is $24.492"
Environment Variables
| Variable | Required | Default | Description |
|---|---|---|---|
NEXUSBRIDGE_API_KEY |
Yes | — | Your NexusBridge API key (starts with nb_sk_) |
NEXUSBRIDGE_BASE_URL |
No | https://syntss.com |
Base URL of the NexusBridge API |
PORT |
No | 8000 |
HTTP port for Smithery/hosted mode |
Transport Modes
| Mode | Flag | Use Case |
|---|---|---|
| HTTP | (default) | Smithery hosted deployment, remote MCP clients |
| stdio | --stdio |
Claude Desktop, Cursor, local MCP clients |
Getting API Credits
| Tier | Credits | Approx. LLM Calls |
|---|---|---|
| Starter | $10 | ~2,000 |
| Builder | $25 | ~5,000 |
| Pro | $50 | ~10,000 |
| Scale | $100 | ~20,000 |
Purchase at syntss.com or via Stripe. Credits never expire.
License
MIT
README copy captured when this server was indexed; it may lag the repository. The original is authoritative.