Koskamo

Cheap GPUs MCP

A read-only MCP server that provides live GPU rental rates, LLM inference pricing, and price forecasts across multiple cloud providers and prediction markets. It allows agents to find the cheapest GPU or LLM route, view price history, and estimate hosting costs.

Visit janus.ai

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.

Cheap GPUs MCP logo, by janus.aiVendor evidenceStreamable HTTPNo sign-inEndpoint checked 2026-11-07Operated by janus.ai
MCP endpoint (Streamable HTTP)
https://cheap-gpus.com/mcp
Vendor documentation

What its tools do

7 tools, grouped by the hints the server advertises. Hints are self-declared, not verified behavior.

Read-only (7)

7 of 7 have a description · 7 of 7 declare behavior hints · 0 of 6 with parameters document them all

Checks and signals

Each line is separate evidence, not a trust score.

  • Passed Endpoint answered the live check ·
  • Passed Official MCP Registry namespace is tied to cheap-gpus.com ·
  • Passed Endpoint is on the registry namespace domain ·
  • Availability 1 of 1 checks passed in 30 days

What you can do with it

  • Find the cheapest GPU rental provider for a specific family
  • Identify the most cost-effective LLM inference route
  • View historical GPU and LLM price trends
  • Estimate GPU requirements and costs for self-hosting models

Source and maintenance

Vendor evidenceRegistry 1.0.0
mcp-registry: com.cheap-gpus/cheap-gpus@1.0.0Project website

Protocol and capabilities

Server
cheap-gpus 1.0.0
MCP protocol negotiated in check
2025-11-25
Tools
7 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.

Live GPU rental prices (USD per GPU-hour) across ~37 cloud providers, LLM inference prices (USD per 1M tokens) across ~24 providers, and market-implied future GPU prices from prediction markets and futures. Prices refresh every 6-8 hours; every row carries synced_at and a stale flag. Use list_gpu_families / find_model to resolve names first. Each result includes a source_url for the matching cheap-gpus.com page.

Tools

7 advertised

list_gpu_families — List GPU families
Advertised by the server during the last successful anonymous check.
Advertised read-only

List every GPU family tracked on cheap-gpus.com, with VRAM, live cheapest on-demand price (USD per GPU-hour), and how many providers currently offer it.

list_gpu_families input schema
{
  "type": "object",
  "title": "list_gpu_familiesArguments",
  "properties": {}
}
list_gpu_families output schema
{
  "type": "object",
  "title": "list_gpu_familiesDictOutput",
  "additionalProperties": true
}
list_gpu_families annotations
{
  "title": "List GPU families",
  "readOnlyHint": true,
  "openWorldHint": false,
  "idempotentHint": true,
  "destructiveHint": false
}
cheapest_gpu — Cheapest GPU rentals
Advertised by the server during the last successful anonymous check.
Advertised read-only

Cheapest current rental offers for one GPU family (e.g. "H100", "RTX 5090", "b200"), in USD per GPU-hour, cheapest first, one row per provider. pricing_type is "on_demand" (default; can't be interrupted) or "spot" (cheaper, the provider can reclaim the machine). limit: 1-10.

cheapest_gpu input schema
{
  "type": "object",
  "title": "cheapest_gpuArguments",
  "required": [
    "gpu"
  ],
  "properties": {
    "gpu": {
      "type": "string",
      "title": "Gpu"
    },
    "limit": {
      "type": "integer",
      "title": "Limit",
      "default": 5
    },
    "pricing_type": {
      "type": "string",
      "title": "Pricing Type",
      "default": "on_demand"
    }
  }
}
cheapest_gpu output schema
{
  "type": "object",
  "title": "cheapest_gpuDictOutput",
  "additionalProperties": true
}
cheapest_gpu annotations
{
  "title": "Cheapest GPU rentals",
  "readOnlyHint": true,
  "openWorldHint": false,
  "idempotentHint": true,
  "destructiveHint": false
}
gpu_price_history — GPU price history
Advertised by the server during the last successful anonymous check.
Advertised read-only

Daily price history for one GPU family: each day's median across providers of each provider's cheapest offer (USD per GPU-hour). days: 1-90.

gpu_price_history input schema
{
  "type": "object",
  "title": "gpu_price_historyArguments",
  "required": [
    "gpu"
  ],
  "properties": {
    "gpu": {
      "type": "string",
      "title": "Gpu"
    },
    "days": {
      "type": "integer",
      "title": "Days",
      "default": 30
    },
    "pricing_type": {
      "type": "string",
      "title": "Pricing Type",
      "default": "on_demand"
    }
  }
}
gpu_price_history output schema
{
  "type": "object",
  "title": "gpu_price_historyDictOutput",
  "additionalProperties": true
}
gpu_price_history annotations
{
  "title": "GPU price history",
  "readOnlyHint": true,
  "openWorldHint": false,
  "idempotentHint": true,
  "destructiveHint": false
}
find_model — Find an LLM
Advertised by the server during the last successful anonymous check.
Advertised read-only

Search the LLM catalog by name (e.g. "deepseek v3", "gpt-oss", "llama 70b"). Returns model slugs to pass to cheapest_model_route / gpus_needed_for_model, most widely offered first. limit: 1-10.

find_model input schema
{
  "type": "object",
  "title": "find_modelArguments",
  "required": [
    "query"
  ],
  "properties": {
    "limit": {
      "type": "integer",
      "title": "Limit",
      "default": 10
    },
    "query": {
      "type": "string",
      "title": "Query"
    }
  }
}
find_model output schema
{
  "type": "object",
  "title": "find_modelDictOutput",
  "additionalProperties": true
}
find_model annotations
{
  "title": "Find an LLM",
  "readOnlyHint": true,
  "openWorldHint": false,
  "idempotentHint": true,
  "destructiveHint": false
}
cheapest_model_route — Cheapest LLM providers
Advertised by the server during the last successful anonymous check.
Advertised read-only

Cheapest inference providers for one LLM (e.g. "deepseek-v3", "gpt-oss-120b"; use find_model if unsure of the name). Prices are USD per 1M tokens; estimated_cost_usd is for the given input/output token counts. One row per provider, cheapest first. limit: 1-10.

cheapest_model_route input schema
{
  "type": "object",
  "title": "cheapest_model_routeArguments",
  "required": [
    "model"
  ],
  "properties": {
    "limit": {
      "type": "integer",
      "title": "Limit",
      "default": 5
    },
    "model": {
      "type": "string",
      "title": "Model"
    },
    "input_tokens": {
      "type": "integer",
      "title": "Input Tokens",
      "default": 1000000
    },
    "output_tokens": {
      "type": "integer",
      "title": "Output Tokens",
      "default": 1000000
    }
  }
}
cheapest_model_route output schema
{
  "type": "object",
  "title": "cheapest_model_routeDictOutput",
  "additionalProperties": true
}
cheapest_model_route annotations
{
  "title": "Cheapest LLM providers",
  "readOnlyHint": true,
  "openWorldHint": false,
  "idempotentHint": true,
  "destructiveHint": false
}
gpu_price_forecast — GPU price forecast
Advertised by the server during the last successful anonymous check.
Advertised read-only

Market-implied future rental price for one GPU family (USD per GPU-hour), from prediction markets (Kalshi, Polymarket) and ICE futures, plus today's compute-index benchmarks and cheap-gpus.com's own current median price. Covers H100, H200, B200, A100, RTX 5090 and RTX PRO 6000.

gpu_price_forecast input schema
{
  "type": "object",
  "title": "gpu_price_forecastArguments",
  "required": [
    "gpu"
  ],
  "properties": {
    "gpu": {
      "type": "string",
      "title": "Gpu"
    }
  }
}
gpu_price_forecast output schema
{
  "type": "object",
  "title": "gpu_price_forecastDictOutput",
  "additionalProperties": true
}
gpu_price_forecast annotations
{
  "title": "GPU price forecast",
  "readOnlyHint": true,
  "openWorldHint": false,
  "idempotentHint": true,
  "destructiveHint": false
}
gpus_needed_for_model — GPUs needed to host a model
Advertised by the server during the last successful anonymous check.
Advertised read-only

How many GPUs it takes to self-host an open-weight LLM at a given quantization, and the cheapest rental setups for it right now. quant is one of: fp16, bf16, fp8, mxfp8, int8, q6, fp6, q5, q4km, nvfp4, awq, mxfp4, gptq, int4, q3, q2, iq2, iq1. Estimates memory only (weights + 20% runtime overhead), not throughput. limit: 1-10.

gpus_needed_for_model input schema
{
  "type": "object",
  "title": "gpus_needed_for_modelArguments",
  "required": [
    "model"
  ],
  "properties": {
    "limit": {
      "type": "integer",
      "title": "Limit",
      "default": 5
    },
    "model": {
      "type": "string",
      "title": "Model"
    },
    "quant": {
      "type": "string",
      "title": "Quant",
      "default": "fp8"
    }
  }
}
gpus_needed_for_model output schema
{
  "type": "object",
  "title": "gpus_needed_for_modelDictOutput",
  "additionalProperties": true
}
gpus_needed_for_model annotations
{
  "title": "GPUs needed to host a model",
  "readOnlyHint": true,
  "openWorldHint": false,
  "idempotentHint": true,
  "destructiveHint": false
}

Prompts

0 advertised

No prompt definitions available.

Resources

0 advertised

No resource definitions available.

90-day uptime

A lightweight connection check about every six hours, sampled on hosts with many listings. It shows the server answered, not that its tools work.

Too few checks for a percentage yet (1 of 1 answered).

  1. 2026-11-07: 1 of 1 checks answered

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
1158 ms
95th percentile latency
1158 ms

Maintenance history

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

  • · probe · passed · 1158 ms · metadata changed · 7 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 cheap-gpus-mcp https://cheap-gpus.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 cheap-gpus-mcp --url https://cheap-gpus.com/mcp

Cursor

Merge this server into .cursor/mcp.json in your project, then enable cheap-gpus-mcp in your MCP settings.
Cursor configuration
{
  "mcpServers": {
    "cheap-gpus-mcp": {
      "url": "https://cheap-gpus.com/mcp"
    }
  }
}

Cline

Merge this into cline_mcp_settings.json (Cline → MCP Servers → Configure).
Cline configuration
{
  "mcpServers": {
    "cheap-gpus-mcp": {
      "type": "streamableHttp",
      "url": "https://cheap-gpus.com/mcp",
      "disabled": false
    }
  }
}

VS Code

Merge this server into .vscode/mcp.json. Use MCP: List Servers to start cheap-gpus-mcp and inspect its tools.
VS Code configuration
{
  "servers": {
    "cheap-gpus-mcp": {
      "type": "http",
      "url": "https://cheap-gpus.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 cheap-gpus-mcp https://cheap-gpus.com/mcp

Claude Desktop

Open Claude Settings > Connectors > Add custom connector, enter this server URL and enable it for the conversation. This is a manual remote-connector handoff, not a desktop JSON configuration.
Claude Desktop configuration
https://cheap-gpus.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": {
    "cheap-gpus-mcp": {
      "serverUrl": "https://cheap-gpus.com/mcp"
    }
  }
}

Let your agent find its tools

Find servers like Cheap GPUs 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 624

Do you run Cheap GPUs MCP?

Prove ownership with a DNS record, a well-known file, or your Official MCP Registry entry, and your edits publish immediately. Proofs are re-checked weekly; admins can still hide or remove a listing.

Claim this listing