Skip to content
MCP server

StudioTV LLM VRAM Calculator

Check whether an LLM fits your GPU, with VRAM, KV cache and GPU counts

Install
Pricing
Free
Platforms
Web, API
Runs
Remotely
Version
1.0.0
Listing updated
1 day ago

About StudioTV LLM VRAM Calculator

StudioTV LLM VRAM Calculator 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.

StudioTV's calculator estimates memory for any Hugging Face model and the MCP endpoint exposes it to assistants. The site's worked example sizes Qwen3.8 27B on one H100 80 GB under vLLM and splits VRAM into weights, KV cache, recurrent state and runtime overhead, accounting for architectures that cache only some layers.

It compares weight formats from BF16 down to FP8, INT8, MXFP4 and INT4 against context lengths from 4K to 256K, and lists the cheapest GPU combinations that fit. Speed and cost are estimated within a stated 30 to 50 percent error margin, set against API prices from OpenRouter and Hugging Face Inference Providers. The site cites a median 1.3 percent gap on weights across 141 real launches.

Rental prices come from 11 providers, checked hourly, and some rental links are affiliate links. The site stresses its figures are estimates and advises confirming with nvidia-smi.

Install

Run in a terminal

claude mcp add --transport http llm-vram https://studiotvai.com/api/mcp

Add to Settings → Connectors → Add custom connector

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

Add to ~/.cursor/mcp.json

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

Add to .vscode/mcp.json

{
  "servers": {
    "llm-vram": {
      "type": "http",
      "url": "https://studiotvai.com/api/mcp"
    }
  }
}

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

{
  "mcpServers": {
    "llm-vram": {
      "serverUrl": "https://studiotvai.com/api/mcp"
    }
  }
}

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

{
  "mcpServers": {
    "llm-vram": {
      "type": "http",
      "url": "https://studiotvai.com/api/mcp"
    }
  }
}

What it does

VRAM breakdown
Weights, KV cache, recurrent state and overhead per GPU.
GPU count
How many GPUs a model needs for given context and concurrency.
Quantization comparison
BF16, FP8, INT8, MXFP4 and INT4 options side by side.
Architecture-aware KV cache
Counts only the layers that cache.
Rental price comparison
Hourly GPU prices across 11 providers.
API cost comparison
Compares self-hosting with API offers.

Good for

  • Check if a model fits a given GPU before downloading
  • Pick the longest context that fits an H100
  • Compare renting GPUs against paying per token
  • Choose a quantization that fits fewer GPUs

Pricing

Free

Free calculator; some rental links are affiliate links

Prices change. Check studiotvai.com before you buy.

StudioTV LLM VRAM Calculator FAQ

Is it free?

The site calls the VRAM calculator free.

How accurate is it?

Memory is within a median 1.3 percent on weights across 141 launches; speed estimates carry a 30 to 50 percent margin.

Which serving engine does it assume?

vLLM in the example, reserving 0.90 GPU memory utilization by default.

Which models work?

Any Hugging Face model, by link or by picking from a list.

Is StudioTV LLM VRAM Calculator free?

Yes. StudioTV LLM VRAM Calculator is free to use.

How do I add StudioTV LLM VRAM Calculator to Claude Code?

Run this command in your terminal: claude mcp add --transport http llm-vram https://studiotvai.com/api/mcp

What are the best alternatives to StudioTV LLM VRAM Calculator?

The closest alternatives to StudioTV LLM VRAM Calculator are Claude Code Ultimate Guide, Agent Yurt, Proofline, signafAIde and Toolkien.

Alternatives to StudioTV LLM VRAM Calculator

  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