# GPUvec

> GPUvec is a free MCP server for cloud and infrastructure. Read-only GPU rental quotes, specs and CUDA data for agents, no API key. It runs as a hosted remote server, so there is nothing to install locally.

- **Kind:** MCP server
- **Pricing:** Free
- **Platforms:** API, Web
- **Website:** https://gpuvec.com/mcp-server
- **MCP Registry name:** io.github.abdelkareemkobo/gpuvec
- **Categories:** cloud and infrastructure, analytics and data
- **Listing updated:** 2026-11-15

## About

Answers carry their sources. The server returns recorded GPU rental quotes with units, provenance and unknown fields kept visible, and the vendor is explicit that this is recorded evidence rather than live inventory, booking or guaranteed final pricing. Price filters work per GPU-hour, with a separate cap for total instance price.

It is a hosted streamable-HTTP endpoint at gpuvec.com/mcp, so nothing runs locally and no account or key is needed. The site gives setup steps for Codex, VS Code and Claude custom connectors, and warns not to use stdio or a legacy SSE endpoint.

Beyond rentals, it covers accelerator specifications, CUDA compute capability and Qdrant sizing. Price history returns only approved observations within inclusive UTC dates, and the site says gaps do not prove a price stayed unchanged.

## Features

- **Offer search:** Filters rental quotes by model, provider, budget, GPU count, VRAM and recorded availability.
- **Spec comparison:** Compares two to four exact catalogue models with explicit units and no invented performance ranking.
- **Price history:** Lists approved price observations between inclusive UTC dates.
- **Qdrant estimates:** Calculates vector storage capacity and cost from stated assumptions, without deploying anything.
- **CUDA lookup:** Returns recorded compute capability, not a guarantee of framework compatibility.
- **Provider directory:** Browses an independent GPU provider list with recorded provenance.

## Good for

- Find H100 quotes under a per-GPU-hour ceiling
- Compare H100 and RTX 4090 specs and offers side by side
- Check CUDA compute capability before choosing a card
- Estimate Qdrant memory and cost for a vector collection
- Review recorded price observations for a date range

## Install

### Claude Code (Terminal)

```sh
claude mcp add --transport http gpuvec https://gpuvec.com/mcp
```

### Claude Desktop (Settings → Connectors → Add custom connector)

```sh
https://gpuvec.com/mcp
```

### Cursor (~/.cursor/mcp.json)

```json
{
  "mcpServers": {
    "gpuvec": {
      "url": "https://gpuvec.com/mcp"
    }
  }
}
```

### VS Code (.vscode/mcp.json)

```json
{
  "servers": {
    "gpuvec": {
      "type": "http",
      "url": "https://gpuvec.com/mcp"
    }
  }
}
```

### Windsurf (~/.codeium/windsurf/mcp_config.json)

```json
{
  "mcpServers": {
    "gpuvec": {
      "serverUrl": "https://gpuvec.com/mcp"
    }
  }
}
```

### Other clients (your client's MCP settings (most use this shape))

```json
{
  "mcpServers": {
    "gpuvec": {
      "type": "http",
      "url": "https://gpuvec.com/mcp"
    }
  }
}
```

## Tools

- `list_accelerators`: Search recorded accelerator models and inspect field-level sources and unknown specifications
- `get_accelerator`: Look up one exact catalogue model name or canonical slug
- `compare_accelerators`: Compare 2-4 exact catalogue models with explicit specification units
- `list_providers`: Browse the independent GPU provider directory and its provenance
- `find_offers`: Filter sourced rental quotes by model, provider, budget, GPU count, VRAM, billing text and availability
- `list_price_history`: Read approved source observations within inclusive UTC capture dates
- `compute_capability`: Look up recorded CUDA compute capability
- `estimate_qdrant`: Calculate vector-storage capacity and estimated cost with explicit assumptions
- `list_qdrant_pricing`: Read dated Qdrant pricing configurations

## FAQ

### Do I need an API key?

No. The endpoint needs no GPUvec account, API key or local install, though your AI client's own plan and policies still apply.

### What is the endpoint URL?

https://gpuvec.com/mcp, used exactly with no trailing slash.

### Are prices live?

No. Results are recorded, sourced quotes, not live stock or guaranteed final pricing.

### Can it change anything?

No. All nine tools are read-only.

### Is GPUvec free?

Yes. GPUvec is free to use.

### How do I add GPUvec to Claude Code?

Run this command in your terminal: claude mcp add --transport http gpuvec https://gpuvec.com/mcp

### What are the best alternatives to GPUvec?

The closest alternatives to GPUvec are Agent-Native Analytics, Google Ads + Meta Ads + SEO/GEO MCP, Tapetide - Indian Stock Market MCP, World Monitor and CompanyData.

## Alternatives

- [Agent-Native Analytics](https://ai-powered.com/mcp/agent-native-analytics): Agent-Native Analytics is an open-source MCP server for analytics and data.
- [Google Ads + Meta Ads + SEO/GEO MCP](https://ai-powered.com/mcp/google-ads-meta-ads-seo-geo-mcp): Google Ads + Meta Ads + SEO/GEO MCP is a freemium MCP server for marketing and seo.
- [Tapetide - Indian Stock Market MCP](https://ai-powered.com/mcp/tapetide-indian-stock-market-mcp): Tapetide - Indian Stock Market MCP is a freemium MCP server for finance and payments.
- [World Monitor](https://ai-powered.com/mcp/world-monitor): World Monitor is a freemium MCP server for analytics and data.
- [CompanyData](https://ai-powered.com/mcp/companydata): CompanyData is a paid MCP server for analytics and data.
- [ChartAnt](https://ai-powered.com/mcp/chartant): ChartAnt is a freemium MCP server for design and media.
- [Quantivex Research](https://ai-powered.com/mcp/quantivex-research): Quantivex Research is a free MCP server for finance and payments.
- [Provyan](https://ai-powered.com/mcp/provyan): Provyan is a free MCP server for finance and payments.

---
Source: https://ai-powered.com/mcp/gpuvec
