Skip to content
MCP server

SpareHive GPU Advisor

Ask which GPU a model needs, whether it fits, and what renting it costs

Install
Pricing
Free
Platforms
API
Runs
Remotely
Version
0.1.0
Listing updated
2 days ago

About SpareHive GPU Advisor

SpareHive GPU Advisor is a free MCP server for ai models and llm tooling. It runs as a hosted remote server, so there is nothing to install locally.

SpareHive's advisor answers from its own dataset and does not use an LLM for the arithmetic. Every number carries an evidence grade. Its example: an RTX 3080 with 10 GB cannot run a 27B model at a 2-bit quantisation, since about 13 GB is needed (weights plus KV cache plus 2 GB overhead), and the server suggests an RTX 4080 16 GB as the first alternative.

Five read-only tools handle model lookup, VRAM requirements per quantisation, fit verdicts (fits, tight, no or unknown) at a given context, GPU rental prices and dataset freshness. It needs no account or key.

The limits are stated plainly: no provisioning, SSH, billing or live inventory, the service is pre-launch, only the RTX 4090 has a published rental price, and image and video models have no KV-cache data so their fit verdicts show unknown.

Install

Run in a terminal

claude mcp add --transport http sparehive-advisor https://sparehive.com/mcp

Add to Settings → Connectors → Add custom connector

https://sparehive.com/mcp
Install in CursorOpens Cursor, which asks you to confirm. Or add it by hand:

Add to ~/.cursor/mcp.json

{
  "mcpServers": {
    "sparehive-advisor": {
      "url": "https://sparehive.com/mcp"
    }
  }
}
Install in VS CodeOpens VS Code, which asks you to confirm. Or add it by hand:

Add to .vscode/mcp.json

{
  "servers": {
    "sparehive-advisor": {
      "type": "http",
      "url": "https://sparehive.com/mcp"
    }
  }
}

Add to ~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "sparehive-advisor": {
      "serverUrl": "https://sparehive.com/mcp"
    }
  }
}

Add to your client's MCP settings (most use this shape)

{
  "mcpServers": {
    "sparehive-advisor": {
      "type": "http",
      "url": "https://sparehive.com/mcp"
    }
  }
}

Tools this server provides (5)

Taken from the project documentation; the running server may differ.

search_models
Finds a model ID by name or alias.
model_requirements
Returns weights and minimum and recommended VRAM for a quantisation.
fit_check
Checks whether a model fits a GPU at a given context.
gpu_prices
Returns rental price, competitor median and market reference per GPU.
data_status
Reports dataset freshness.

What it does

Model search
search_models finds a model ID by name or alias and type.
VRAM requirements
model_requirements gives weights and minimum and recommended VRAM.
Fit checks
fit_check returns fits, tight, no or unknown with headroom.
Price comparison
gpu_prices shows rental and competitor median prices.
Freshness check
data_status flags files older than 60 days.

Good for

  • Check whether a local model fits a GPU you own
  • Estimate VRAM for a chosen quantisation
  • Find a cheaper card as an alternative
  • Compare rental cost with market references

Pricing

Free

Free, read-only, no account

Prices change. Check sparehive.com before you buy.

SpareHive GPU Advisor FAQ

Does it need an account?

No account or API key is needed.

Can it rent or provision a GPU?

No. It has no provisioning, SSH, billing or live inventory.

Which GPUs have a rental price?

Only the RTX 4090; others show not priced yet.

Why does an image model show unknown?

Image and video models have no KV-cache data in the dataset.

Is SpareHive GPU Advisor free?

Yes. SpareHive GPU Advisor is free to use.

How do I add SpareHive GPU Advisor to Claude Code?

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

What are the best alternatives to SpareHive GPU Advisor?

The closest alternatives to SpareHive GPU Advisor are Claude Code Ultimate Guide, Agent Yurt, Proofline, signafAIde and Toolkien.

Alternatives to SpareHive GPU Advisor

  1. MCP server for searching the Claude Code Ultimate Guide from any MCP client

    Runs locally
    • Open source
    • 6.5K GitHub stars
    • AI models and LLM tooling
  2. Agent Yurt

    MCP server

    Back up agent memory, files and dashboards, then recover from a reviewed snapshot

    Not checked yet
    • Remote
    • Free
    • Knowledge and memory
  3. Proofline

    MCP server

    Production-aware review of changes against your AWS or Google Cloud setup, via MCP

    • Remote
    • Freemium
    • Developer tools
  4. signafAIde

    MCP server

    Local MCP server to set up, verify and check coverage of signed agent action records

    Runs locally
    • 5 tools
    • Freemium
    • Security
  5. Toolkien

    MCP server

    One MCP gateway for all your MCP servers, with code mode, permissions and audit

    Not checked yet
    • 3 tools
    • Remote
    • Open source
    • AI models and LLM tooling
  6. filex

    MCP server

    Self-hosted file manager in one Go binary with a built-in MCP server for agents

    Not checked yet
    • Remote
    • Open source
    • Files and documents