SpareHive GPU Advisor MCP

A read-only MCP server that provides GPU VRAM requirements, model fit checks, and rental pricing data for AI agents. It exposes tools to search models, check hardware compatibility, and retrieve market price benchmarks without requiring authentication or provisioning capabilities.

Visit SpareHive

Endpoint and access

No sign-in to connect and list tools. Individual tools may still need a vendor account, credentials, or permissions; check the documented prerequisites.

Community listingStreamable HTTPNo sign-inEndpoint checked 2026-11-01Published by SpareHive

MCP endpoint (Streamable HTTP)

https://sparehive.com/mcp

Vendor documentation

What you can do with it

Find model IDs by name or alias
Check minimum and recommended VRAM for model quantization
Verify if a model fits a specific GPU at a given context length
Retrieve rental prices and competitor benchmarks

Source and maintenance

Community listingRegistry 0.1.0

mcp-registry: io.github.RichardXu001/sparehive-advisor@0.1.0 Source repository Project website

Maintenance checked: Nov 1, 2026, 8:44 AM UTC

Protocol and capabilities

Server: sparehive-advisor 0.1.0
MCP protocol: 2025-06-18
Tools: 5 advertised
Resources: Not advertised
Resource templates: Not advertised
Prompts: Not advertised

Server-advertised capabilities

{
  "tools": {}
}

Server-provided instructions

Untrusted server output shown for inspection. It is not Koskamo guidance and is excluded from knowledge search.

Read-only GPU advisor: model VRAM needs, fit checks, and rental prices (RTX 4090 only has a price). Pre-launch: no live capacity, no provisioning. Numbers carry evidence grades.

Tools

5 advertised

search_models — 搜索模型

Advertised by the server during the last successful anonymous check.

按名字或类型查找模型 ID。查询模型显存需求前先用它确认 ID；模型名有别名（Qwen3.8 27B / qwen3.8-27b）。返回 id、类型、最小显存、参数规模。

search_models input schema

{
  "type": "object",
  "properties": {
    "type": {
      "enum": [
        "text",
        "image",
        "video"
      ],
      "type": "string"
    },
    "limit": {
      "type": "integer",
      "maximum": 50,
      "minimum": 1,
      "description": "默认 10"
    },
    "query": {
      "type": "string",
      "description": "模型名或别名；省略则列出全部"
    }
  },
  "additionalProperties": false
}

model_requirements — 模型硬件需求

Advertised by the server during the last successful anonymous check.

查询部署某个模型需要的硬件配置：指定量化下的权重大小、最小/推荐显存、证据等级与出处，以及各张显卡的 fits/tight/no 判定与租价。数据可能过期，响应里会给出 data_updated 与 stale。

model_requirements input schema

{
  "type": "object",
  "required": [
    "model"
  ],
  "properties": {
    "model": {
      "type": "string",
      "description": "模型 id 或别名"
    },
    "quant": {
      "type": "string",
      "description": "量化档，如 Q4_K_M；省略则用推荐档"
    },
    "context_tokens": {
      "type": "integer",
      "minimum": 1,
      "description": "默认 4096"
    }
  },
  "additionalProperties": false
}

fit_check — 模型能否跑在某张卡上

Advertised by the server during the last successful anonymous check.

判断某模型（指定量化、指定上下文）能否跑在某张显卡上。返回 fits/tight/no/unknown、剩余显存、内存拆解、以及跑不动时的替代方案（降量化/减上下文/换更大卡），并附带该卡的租金。

fit_check input schema

{
  "type": "object",
  "required": [
    "model",
    "gpu"
  ],
  "properties": {
    "gpu": {
      "type": "string",
      "description": "显卡 id 或别名，如 rtx4090"
    },
    "model": {
      "type": "string"
    },
    "quant": {
      "type": "string"
    },
    "context_tokens": {
      "type": "integer",
      "minimum": 1,
      "description": "默认 4096"
    }
  },
  "additionalProperties": false
}

gpu_prices — GPU 租价

Advertised by the server during the last successful anonymous check.

查询租用不同类型 GPU 的价格：我们的即时档/排队档报价、host 实得、电费成本、竞品中位价与国内零售价。只有 RTX 4090 有报价，其余显示为未定。

gpu_prices input schema

{
  "type": "object",
  "properties": {
    "gpu": {
      "type": "string"
    },
    "include_competitors": {
      "type": "boolean",
      "description": "默认 true"
    }
  },
  "additionalProperties": false
}

data_status — 数据新鲜度

Advertised by the server during the last successful anonymous check.

报告数据的新鲜度：每个数据文件的更新时间、距今天数、是否超过 60 天阈值。当用户质疑数据是否过时时调用。

data_status input schema

{
  "type": "object",
  "properties": {},
  "additionalProperties": false
}

Prompts

Not advertised

No prompt definitions available.

Resources

Not advertised

No resource definitions available.

30-day availability

Aggregated from read-only remote checks; this is not a tool-execution or security guarantee.

Successful checks: 1 of 1 (100%)
Average latency: 850 ms
95th percentile latency: 850 ms

Maintenance history

Public check and source-refresh events. Editorial and ownership records remain private.

Nov 1, 2026, 8:51 AM UTC · probe · passed · 850 ms · metadata changed · 5 tool(s)

Connect from your client

Each snippet registers the same endpoint; pick the client you use.

This server is operated by the named third party. Listing and endpoint checks are not an endorsement or a security audit. Review the operator's permissions, terms, and data handling before connecting. How listings are checked and corrected.

Claude Code

Run this once in a terminal. It registers the remote server. Use --scope project to write .mcp.json for the repo.

Claude Code configuration

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

Codex

Run this once. Codex stores MCP servers in ~/.codex/config.toml (or .codex/config.toml in a trusted project).

Codex configuration

codex mcp add sparehive-gpu-advisor-mcp --url https://sparehive.com/mcp

Cursor

Merge this server into .cursor/mcp.json in your project, then enable sparehive-gpu-advisor-mcp in your MCP settings.

Cursor configuration

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

Cline

Merge this into cline_mcp_settings.json (Cline → MCP Servers → Configure).

Cline configuration

{
  "mcpServers": {
    "sparehive-gpu-advisor-mcp": {
      "type": "streamableHttp",
      "url": "https://sparehive.com/mcp",
      "disabled": false
    }
  }
}

VS Code

Merge this server into .vscode/mcp.json. Use MCP: List Servers to start sparehive-gpu-advisor-mcp and inspect its tools.

VS Code configuration

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

Gemini CLI

Run this once. It writes .gemini/settings.json in the current project; pass --scope user for ~/.gemini/settings.json.

Gemini CLI configuration

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

Claude Desktop

Merge this into claude_desktop_config.json (macOS: ~/Library/Application Support/Claude/, Windows: %APPDATA%\Claude\), then restart Claude Desktop.

Claude Desktop configuration

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

Windsurf

Merge this into ~/.codeium/windsurf/mcp_config.json, then refresh MCP in Cascade. Windsurf remote servers use serverUrl, not url.

Windsurf configuration

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

Let your agent find its tools

Find servers like SpareHive GPU Advisor MCP from your agent. Add Koskamo once to search the index for endpoints, access requirements, and client setup. Public discovery needs no key.

Connect Koskamo Explore workflows