tokcalc MCP Server
An open-source capacity planner for LLM deployments that calculates VRAM usage, KV cache sizing, throughput, latency (TTFT/ITL), and cost across various models, GPUs, and quantizations. It supports CI/CD integration via GitHub Actions and local CLI execution to help users plan GPU resources before renting.
Package and access
Local package with source-attributed installation metadata. Koskamo does not execute packages during review.
What you can do with it
- Calculate VRAM budget and KV-cache sizing for specific models and GPUs
- Estimate throughput and latency metrics like TTFT and ITL
- Compare build-vs-buy costs across cloud providers and self-hosting
- Integrate capacity planning checks into CI/CD pipelines via GitHub Actions
Source and maintenance
- Maintenance checked
Packages and installation
Source-provided package data. Koskamo has not executed or live-probed this package.
@tokcalc/mcp-server 0.3.4
{
"version": "0.3.4",
"transport": "stdio",
"identifier": "@tokcalc/mcp-server",
"registryType": "npm"
}Tools
Reported by the attributed source; not independently observed by Koskamo.
Connect from your client
Each snippet configures the same source-provided package; replace placeholders before use.
Cursor
{
"mcpServers": {
"tokcalc-mcp-server": {
"command": "npx",
"args": [
"-y",
"@tokcalc/mcp-server@0.3.4"
]
}
}
}Cline
{
"mcpServers": {
"tokcalc-mcp-server": {
"command": "npx",
"args": [
"-y",
"@tokcalc/mcp-server@0.3.4"
]
}
}
}VS Code
{
"servers": {
"tokcalc-mcp-server": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@tokcalc/mcp-server@0.3.4"
]
}
}
}Claude Desktop
{
"mcpServers": {
"tokcalc-mcp-server": {
"command": "npx",
"args": [
"-y",
"@tokcalc/mcp-server@0.3.4"
]
}
}
}Let your agent find its tools
Find servers like tokcalc MCP Server from your agent. Add Koskamo once to search the index for endpoints, access requirements, and client setup. Public discovery needs no key.