GPUvec

io.github.abdelkareemkoboPublished by GitHub account abdelkareemkobov1.0.2Updated Nov 1, 2026

Compare GPU specs and cloud rental offers, check CUDA support, and estimate Qdrant resources.

VerifiedStreamable HTTPWeb executableCloud & InfrastructureData & Analytics

Overview

AI-generated overview

Read-only public catalogue for GPU specs, cloud GPU rental offers, CUDA compute capability and Qdrant capacity estimates.

What you can do
  • Look up accelerator specifications, including units, field provenance and unknowns, and fetch one exact model by name or slug
  • Compare two to four exact accelerator models side by side without invented scores or rankings
  • List GPU compute providers with official provenance, and find recorded rental offers filtered by price, count, VRAM and billing
  • Read approved GPU price observations newest first, filtered by UTC capture date
  • Retrieve source-backed NVIDIA CUDA compute capability context for a model
  • Estimate Qdrant capacity with the site's shared engine and list dated official managed Cloud package quotes
When to use it
Useful when you need to compare GPU hardware or cloud rental offers, check CUDA compute capability, or size a Qdrant deployment while keeping sources and unknowns explicit. It suits research and planning questions where recorded quotes and provenance matter more than live availability.
Requirements
Remote streamable HTTP endpoint at gpuvec.com; no package, account, API key or environment variable is declared. The assistant needs network access to that endpoint.
Before you install
Prices and availability are recorded quotes, not live stock, and Qdrant estimates are deterministic calculations rather than measured capacity or QPS sizing. Unknown fields are reported as unknown rather than zero, and compute capability does not guarantee toolkit or framework support. The server is described as read-only and as not fetching external data or private data.

Installation

In SourceWeft

  1. Open GPUvec in the dashboard and add it to a workspace.
  2. Enable the server for the chats that should use its tools.

Web executable via Streamable HTTP. Remote servers run from the web runtime once configured in a workspace.

Use with an agent

Paste this prompt into Claude Code, Codex, Cursor or another agent — or into SourceWeft chat. The agent reads this server's install guide, shows you its publisher, endpoint and sign-in, and adds it to your client once you agree. Keys never go into a command.

Read https://sourceweft.com/mcp/io.github.abdelkareemkobo%2Fgpuvec/install.md and connect the MCP server it describes to my client. Before changing any config, show me its publisher, endpoint and sign-in method and wait for my OK. Never put an API key in a command or in a file I would commit; ask me where it should come from.

Read the install guide the agent follows

Other MCP clients

Add this to your client's mcpServers config.

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

Tools

9

list_accelerators

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

Parameterslimitoffsetsearchmanufacturer

get_accelerator

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

Parametersmodelrequired

compare_accelerators

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

Parametersmodelsrequired

list_providers

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

Parameterslimitoffsetsearch

find_offers

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.

Parameterssortlimitmodeloffsetprovideravailabilitybilling_modelmin_gpu_countmin_vram_totalmin_vram_per_gpusource_backed_onlymax_price_per_gpu_hour+1 more

list_price_history

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

Parametersendlimitmodelstartoffsetprovider

compute_capability

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

Parameterslimitoffsetsearchstatussegmentcapability

estimate_qdrant

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

Parametersinputsrequired

list_qdrant_pricing

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

Parameterslimitoffsetregionprovider

Version history

1
  1. v1.0.2LatestNov 1, 2026

Score

83% of the applicable checks met

How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.

  • Reached in the last 30 daysRequired25 pts
  • Auth declaredRequired10 pts
  • Tools listedRequired15 pts
  • DescriptionRequired15 pts
  • Logo10 pts
  • Documentation15 pts

    Declare a GitHub repository with a README, or serve resource_documentation in the OAuth protected-resource metadata.

  • LicenseNot applicable

    A hosted service is used under its provider's terms, so no license is expected here. This check is left out of the percentage.