GPUvec MCP

A public, read-only MCP server that connects AI agents to GPUvec's database of sourced GPU rental quotes and accelerator specifications. It provides tools to compare GPU models, find cloud offers within budget constraints, and estimate resources for vector databases like Qdrant, without requiring authentication or local installation.

Visit GPUvec

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 GPUvec

MCP endpoint (Streamable HTTP)

https://gpuvec.com/mcp

Vendor documentation

What you can do with it

Find H100 quotes within a specific budget and memory requirements
Compare exact configurations and normalized prices between GPU models
Check CUDA support and compute capability for specific accelerators
Estimate Qdrant vector database resource requirements

Source and maintenance

Community listingRegistry 1.0.2

mcp-registry: io.github.abdelkareemkobo/gpuvec@1.0.2 Project website

Protocol and capabilities

Server: GPUvec 1.0.0
MCP protocol: 2025-06-18
Tools: 9 advertised
Resources: 0 advertised
Resource templates: 0 advertised
Prompts: 0 advertised

Server-advertised capabilities

{
  "tools": {
    "listChanged": false
  },
  "prompts": {
    "listChanged": false
  },
  "resources": {
    "subscribe": false,
    "listChanged": false
  }
}

Server-provided instructions

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

Read-only public catalogue. Cite canonical URLs and source provenance. Unknown fields are not zero. Prices are recorded quotes, not live stock. No external fetch or private data.

Tools

9 advertised

list_accelerators

Advertised by the server during the last successful anonymous check.

Advertised read-only

List recorded accelerator specifications, with units, field provenance and unknowns. At most 50 per page.

list_accelerators input schema

{
  "type": "object",
  "title": "list_acceleratorsArguments",
  "properties": {
    "limit": {
      "type": "integer",
      "title": "Limit",
      "default": 20,
      "maximum": 50,
      "minimum": 1
    },
    "offset": {
      "type": "integer",
      "title": "Offset",
      "default": 0,
      "maximum": 10000,
      "minimum": 0
    },
    "search": {
      "type": "string",
      "title": "Search",
      "default": "",
      "pattern": "^[^\\x00-\\x1f\\x7f]*$",
      "maxLength": 120
    },
    "manufacturer": {
      "type": "string",
      "title": "Manufacturer",
      "default": "",
      "pattern": "^[^\\x00-\\x1f\\x7f]*$",
      "maxLength": 120
    }
  }
}

list_accelerators output schema

{
  "type": "object",
  "title": "list_acceleratorsDictOutput",
  "additionalProperties": true
}

list_accelerators annotations

{
  "readOnlyHint": true,
  "openWorldHint": false
}

get_accelerator

Advertised by the server during the last successful anonymous check.

Advertised read-only

Get one exact accelerator model or canonical slug. Unknown models are errors, not guessed specifications.

get_accelerator input schema

{
  "type": "object",
  "title": "get_acceleratorArguments",
  "required": [
    "model"
  ],
  "properties": {
    "model": {
      "type": "string",
      "title": "Model",
      "pattern": "^[^\\x00-\\x1f\\x7f]+$",
      "maxLength": 120,
      "minLength": 1
    }
  }
}

get_accelerator output schema

{
  "type": "object",
  "title": "get_acceleratorDictOutput",
  "additionalProperties": true
}

get_accelerator annotations

{
  "readOnlyHint": true,
  "openWorldHint": false
}

compare_accelerators

Advertised by the server during the last successful anonymous check.

Advertised read-only

Compare 2–4 distinct exact models side by side; no invented performance score or compatibility ranking.

compare_accelerators input schema

{
  "type": "object",
  "title": "compare_acceleratorsArguments",
  "required": [
    "models"
  ],
  "properties": {
    "models": {
      "type": "array",
      "items": {
        "type": "string",
        "pattern": "^[^\\x00-\\x1f\\x7f]+$",
        "maxLength": 120,
        "minLength": 1
      },
      "title": "Models",
      "maxItems": 4,
      "minItems": 2
    }
  }
}

compare_accelerators output schema

{
  "type": "object",
  "title": "compare_acceleratorsDictOutput",
  "additionalProperties": true
}

compare_accelerators annotations

{
  "readOnlyHint": true,
  "openWorldHint": false
}

list_providers

Advertised by the server during the last successful anonymous check.

Advertised read-only

List GPU compute services with field-scoped official provenance, independent of priced offers and live stock.

list_providers input schema

{
  "type": "object",
  "title": "list_providersArguments",
  "properties": {
    "limit": {
      "type": "integer",
      "title": "Limit",
      "default": 20,
      "maximum": 50,
      "minimum": 1
    },
    "offset": {
      "type": "integer",
      "title": "Offset",
      "default": 0,
      "maximum": 10000,
      "minimum": 0
    },
    "search": {
      "type": "string",
      "title": "Search",
      "default": "",
      "pattern": "^[^\\x00-\\x1f\\x7f]*$",
      "maxLength": 120
    }
  }
}

list_providers output schema

{
  "type": "object",
  "title": "list_providersDictOutput",
  "additionalProperties": true
}

list_providers annotations

{
  "readOnlyHint": true,
  "openWorldHint": false
}

find_offers

Advertised by the server during the last successful anonymous check.

Advertised read-only

Find recorded offers with bounded price/count/VRAM filters. Unknown units, scope or counts never pass numeric filters. Billing is a case-insensitive substring; availability is recorded, not live.

find_offers input schema

{
  "type": "object",
  "title": "find_offersArguments",
  "properties": {
    "sort": {
      "enum": [
        "price_per_gpu_hour",
        "-price_per_gpu_hour",
        "instance_price_per_hour",
        "-instance_price_per_hour"
      ],
      "type": "string",
      "title": "Sort",
      "default": "price_per_gpu_hour"
    },
    "limit": {
      "type": "integer",
      "title": "Limit",
      "default": 20,
      "maximum": 50,
      "minimum": 1
    },
    "model": {
      "type": "string",
      "title": "Model",
      "default": "",
      "pattern": "^[^\\x00-\\x1f\\x7f]*$",
      "maxLength": 120
    },
    "offset": {
      "type": "integer",
      "title": "Offset",
      "default": 0,
      "maximum": 10000,
      "minimum": 0
    },
    "provider": {
      "type": "string",
      "title": "Provider",
      "default": "",
      "pattern": "^[^\\x00-\\x1f\\x7f]*$",
      "maxLength": 120
    },
    "availability": {
      "enum": [
        "",
        "available",
        "unavailable",
        "waitlist",
        "coming_soon",
        "unknown"
      ],
      "type": "string",
      "title": "Availability",
      "default": ""
    },
    "billing_model": {
      "type": "string",
      "title": "Billing Model",
      "default": "",
      "pattern": "^[^\\x00-\\x1f\\x7f]*$",
      "maxLength": 120
    },
    "min_gpu_count": {
      "anyOf": [
        {
          "type": "integer",
          "maximum": 1000000,
          "minimum": 1
        },
        {
          "type": "null"
        }
      ],
      "title": "Min Gpu Count",
      "default": null
    },
    "min_vram_total": {
      "anyOf": [
        {
          "type": "number",
          "maximum": 1000000000,
          "minimum": 0
        },
        {
          "type": "null"
        }
      ],
      "title": "Min Vram Total",
      "default": null
    },
    "min_vram_per_gpu": {
      "anyOf": [
        {
          "type": "number",
          "maximum": 1000000000,
          "minimum": 0
        },
        {
          "type": "null"
        }
      ],
      "title": "Min Vram Per Gpu",
      "default": null
    },
    "source_backed_only": {
      "type": "boolean",
      "title": "Source Backed Only",
      "default": true
    },
    "max_price_per_gpu_hour": {
      "anyOf": [
        {
          "type": "number",
          "maximum": 1000000000,
          "minimum": 0
        },
        {
          "type": "null"
        }
      ],
      "title": "Max Price Per Gpu Hour",
      "default": null
    },
    "max_instance_price_per_hour": {
      "anyOf": [
        {
          "type": "number",
          "maximum": 1000000000,
          "minimum": 0
        },
        {
          "type": "null"
        }
      ],
      "title": "Max Instance Price Per Hour",
      "default": null
    }
  }
}

find_offers output schema

{
  "type": "object",
  "title": "find_offersDictOutput",
  "additionalProperties": true
}

find_offers annotations

{
  "readOnlyHint": true,
  "openWorldHint": false
}

list_price_history

Advertised by the server during the last successful anonymous check.

Advertised read-only

Read actual approved GPU price observations, newest first. UTC dates filter capture time; distinct configurations are not one price trend.

list_price_history input schema

{
  "type": "object",
  "title": "list_price_historyArguments",
  "properties": {
    "end": {
      "type": "string",
      "title": "End",
      "default": "",
      "pattern": "^(?:\\d{4}-\\d{2}-\\d{2})?$",
      "maxLength": 10
    },
    "limit": {
      "type": "integer",
      "title": "Limit",
      "default": 20,
      "maximum": 50,
      "minimum": 1
    },
    "model": {
      "type": "string",
      "title": "Model",
      "default": "",
      "pattern": "^[^\\x00-\\x1f\\x7f]*$",
      "maxLength": 120
    },
    "start": {
      "type": "string",
      "title": "Start",
      "default": "",
      "pattern": "^(?:\\d{4}-\\d{2}-\\d{2})?$",
      "maxLength": 10
    },
    "offset": {
      "type": "integer",
      "title": "Offset",
      "default": 0,
      "maximum": 10000,
      "minimum": 0
    },
    "provider": {
      "type": "string",
      "title": "Provider",
      "default": "",
      "pattern": "^[^\\x00-\\x1f\\x7f]*$",
      "maxLength": 120
    }
  }
}

list_price_history output schema

{
  "type": "object",
  "title": "list_price_historyDictOutput",
  "additionalProperties": true
}

list_price_history annotations

{
  "readOnlyHint": true,
  "openWorldHint": false
}

compute_capability

Advertised by the server during the last successful anonymous check.

Advertised read-only

Retrieve source-backed NVIDIA CUDA hardware compute capability context; non-CUDA is not assigned a CC. No toolkit/framework guarantee.

compute_capability input schema

{
  "type": "object",
  "title": "compute_capabilityArguments",
  "properties": {
    "limit": {
      "type": "integer",
      "title": "Limit",
      "default": 20,
      "maximum": 50,
      "minimum": 1
    },
    "offset": {
      "type": "integer",
      "title": "Offset",
      "default": 0,
      "maximum": 10000,
      "minimum": 0
    },
    "search": {
      "type": "string",
      "title": "Search",
      "default": "",
      "pattern": "^[^\\x00-\\x1f\\x7f]*$",
      "maxLength": 120
    },
    "status": {
      "enum": [
        "",
        "known",
        "unknown",
        "not_cuda"
      ],
      "type": "string",
      "title": "Status",
      "default": ""
    },
    "segment": {
      "enum": [
        "",
        "data_center",
        "workstation",
        "jetson",
        "other"
      ],
      "type": "string",
      "title": "Segment",
      "default": ""
    },
    "capability": {
      "type": "string",
      "title": "Capability",
      "default": "",
      "pattern": "^(?:\\d{1,2}\\.\\d{1,2})?$",
      "maxLength": 6
    }
  }
}

compute_capability output schema

{
  "type": "object",
  "title": "compute_capabilityDictOutput",
  "additionalProperties": true
}

compute_capability annotations

{
  "readOnlyHint": true,
  "openWorldHint": false
}

estimate_qdrant

Advertised by the server during the last successful anonymous check.

Advertised read-only

Calculate deterministic Qdrant capacity using the site's shared real engine. KiB inputs, byte outputs; not measured capacity or QPS sizing.

estimate_qdrant input schema

{
  "type": "object",
  "$defs": {
    "QdrantInputs": {
      "type": "object",
      "title": "QdrantInputs",
      "properties": {
        "mmap": {
          "type": "boolean",
          "title": "Mmap",
          "default": false
        },
        "nodes": {
          "type": "integer",
          "title": "Nodes",
          "default": 1,
          "maximum": 100000,
          "minimum": 1
        },
        "hnsw_m": {
          "type": "integer",
          "title": "Hnsw M",
          "default": 16,
          "maximum": 10000,
          "minimum": 0
        },
        "shards": {
          "type": "integer",
          "title": "Shards",
          "default": 1,
          "maximum": 1000000,
          "minimum": 1
        },
        "datatype": {
          "enum": [
            "float32",
            "float16",
            "uint8",
            "turbo16",
            "turbo8",
            "turbo4"
          ],
          "type": "string",
          "title": "Datatype",
          "default": "float32"
        },
        "dimensions": {
          "type": "integer",
          "title": "Dimensions",
          "default": 1536,
          "maximum": 65536,
          "minimum": 0
        },
        "num_points": {
          "type": "integer",
          "title": "Num Points",
          "default": 1000000,
          "maximum": 1000000000000,
          "minimum": 0
        },
        "num_sparse": {
          "type": "integer",
          "title": "Num Sparse",
          "default": 0,
          "maximum": 1000000000000,
          "minimum": 0
        },
        "payload_kb": {
          "type": "number",
          "title": "Payload Kb",
          "default": 1,
          "maximum": 1000000000,
          "minimum": 0,
          "description": "KiB per unique point, not decimal kB"
        },
        "num_vectors": {
          "type": "integer",
          "title": "Num Vectors",
          "default": 1000000,
          "maximum": 1000000000000,
          "minimum": 0
        },
        "replication": {
          "type": "integer",
          "title": "Replication",
          "default": 1,
          "maximum": 100000,
          "minimum": 1
        },
        "hnsw_on_disk": {
          "type": "boolean",
          "title": "Hnsw On Disk",
          "default": false
        },
        "quantization": {
          "enum": [
            "none",
            "scalar",
            "binary",
            "binary1_5",
            "binary2",
            "product",
            "turbo"
          ],
          "type": "string",
          "title": "Quantization",
          "default": "none"
        },
        "billing_hours": {
          "type": "number",
          "title": "Billing Hours",
          "default": 730,
          "maximum": 8784,
          "minimum": 0
        },
        "cache_percent": {
          "type": "number",
          "title": "Cache Percent",
          "default": 0,
          "maximum": 100,
          "minimum": 0
        },
        "sparse_on_disk": {
          "type": "boolean",
          "title": "Sparse On Disk",
          "default": false
        },
        "hourly_per_node": {
          "anyOf": [
            {
              "type": "number",
              "maximum": 1000000000,
              "minimum": 0,
              "description": "User-supplied USD/node-hour; null means unknown"
            },
            {
              "type": "null"
            }
          ],
          "title": "Hourly Per Node",
          "default": null
        },
        "payload_on_disk": {
          "type": "boolean",
          "title": "Payload On Disk",
          "default": true
        },
        "sparse_elements": {
          "type": "number",
          "title": "Sparse Elements",
          "default": 0,
          "maximum": 4294967295,
          "minimum": 0
        },
        "headroom_percent": {
          "type": "number",
          "title": "Headroom Percent",
          "default": 20,
          "maximum": 1000,
          "minimum": 0
        },
        "pricing_quote_id": {
          "type": "string",
          "title": "Pricing Quote Id",
          "default": "custom",
          "pattern": "^[^\\x00-\\x1f\\x7f]*$",
          "maxLength": 120
        },
        "quantization_bits": {
          "enum": [
            "bits8",
            "bits4",
            "bits2",
            "bits1_5",
            "bits1"
          ],
          "type": "string",
          "title": "Quantization Bits",
          "default": "bits4"
        },
        "indexed_payload_kb": {
          "type": "number",
          "title": "Indexed Payload Kb",
          "default": 0,
          "maximum": 1000000000,
          "minimum": 0,
          "description": "KiB per unique point"
        },
        "product_compression": {
          "enum": [
            4,
            8,
            16,
            32,
            64
          ],
          "type": "integer",
          "title": "Product Compression",
          "default": 8
        },
        "quantization_memory": {
          "enum": [
            "auto",
            "pinned",
            "cached",
            "cold"
          ],
          "type": "string",
          "title": "Quantization Memory",
          "default": "auto"
        },
        "sparse_bytes_per_nonzero": {
          "type": "number",
          "title": "Sparse Bytes Per Nonzero",
          "default": 8,
          "maximum": 1024,
          "minimum": 0.001
        }
      },
      "description": "Same field names, units and ranges as the site's shared Qdrant engine.",
      "additionalProperties": false
    }
  },
  "title": "estimate_qdrantArguments",
  "required": [
    "inputs"
  ],
  "properties": {
    "inputs": {
      "$ref": "#/$defs/QdrantInputs"
    }
  }
}

estimate_qdrant output schema

{
  "type": "object",
  "title": "estimate_qdrantDictOutput",
  "additionalProperties": true
}

estimate_qdrant annotations

{
  "readOnlyHint": true,
  "openWorldHint": false
}

list_qdrant_pricing

Advertised by the server during the last successful anonymous check.

Advertised read-only

List dated official managed Cloud package quotes, not live prices or hardware recommendations.

list_qdrant_pricing input schema

{
  "type": "object",
  "title": "list_qdrant_pricingArguments",
  "properties": {
    "limit": {
      "type": "integer",
      "title": "Limit",
      "default": 20,
      "maximum": 50,
      "minimum": 1
    },
    "offset": {
      "type": "integer",
      "title": "Offset",
      "default": 0,
      "maximum": 10000,
      "minimum": 0
    },
    "region": {
      "type": "string",
      "title": "Region",
      "default": "",
      "pattern": "^[^\\x00-\\x1f\\x7f]*$",
      "maxLength": 120
    },
    "provider": {
      "type": "string",
      "title": "Provider",
      "default": "",
      "pattern": "^[^\\x00-\\x1f\\x7f]*$",
      "maxLength": 120
    }
  }
}

list_qdrant_pricing output schema

{
  "type": "object",
  "title": "list_qdrant_pricingDictOutput",
  "additionalProperties": true
}

list_qdrant_pricing annotations

{
  "readOnlyHint": true,
  "openWorldHint": false
}

Prompts

0 advertised

No prompt definitions available.

Resources

0 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: 536 ms
95th percentile latency: 536 ms

Maintenance history

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

Nov 1, 2026, 7:11 PM UTC · probe · passed · 536 ms · metadata changed · 9 tool(s), 0 resource(s), 0 resource template(s), 0 prompt(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 gpuvec-mcp https://gpuvec.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 gpuvec-mcp --url https://gpuvec.com/mcp

Cursor

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

Cursor configuration

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

Cline

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

Cline configuration

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

VS Code

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

VS Code configuration

{
  "servers": {
    "gpuvec-mcp": {
      "type": "http",
      "url": "https://gpuvec.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 gpuvec-mcp https://gpuvec.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": {
    "gpuvec-mcp": {
      "url": "https://gpuvec.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": {
    "gpuvec-mcp": {
      "serverUrl": "https://gpuvec.com/mcp"
    }
  }
}

Let your agent find its tools

Find servers like GPUvec 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

See all 538

GPUvec MCP

Endpoint and access

What you can do with it

Source and maintenance

Protocol and capabilities

Server-provided instructions

Tools

Prompts

Resources

30-day availability

Maintenance history

Connect from your client

Claude Code

Codex

Cursor

Cline

VS Code

Gemini CLI

Claude Desktop

Windsurf

Let your agent find its tools

Other cloud and infrastructure servers