TokenMark

io.github.sypherinv0.2.1Updated Oct 5, 2026

Measured local-LLM speeds and hardware-aware model picks for Strix Halo, DGX Spark and Mac.

VerifiedSTDIODesktop onlyAI & MLData & Analytics

Overview

AI-generated overview

Lets an assistant look up measured local-LLM benchmark speeds and hardware-aware model recommendations from TokenMark.

What it does
Provides tools to recommend ranked model and configuration picks for a given hardware platform and task, list tracked benchmark configurations with measured decode tokens per second, browse a hardware catalogue with chip specs and prices, search models and configs, and submit a repository with benchmark numbers for human review. Data is fetched live from the TokenMark site, and the server claims it never invents numbers.
When to use it
Useful when choosing which local model or quantization to run on specific hardware such as Strix Halo, DGX Spark or Mac, or when comparing measured throughput figures instead of estimates.
Requirements
Runs locally over stdio, typically launched with npx, so Node.js and network access to the TokenMark site are needed. No accounts, API keys or environment variables are declared; the data source URL can be overridden with the TOKENMARK_URL environment variable.
Before you install
The server reads live data from a third-party site, so queries and any submitted repository details leave the machine. The submit tool queues third-party repository links and notes for human review, and the README points to a signed-in web form for personal measurements.

Installation

In SourceWeft

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

Desktop only via STDIO. STDIO servers start a local process, so they need the SourceWeft desktop host.

Other MCP clients

Follow the launch instructions in the repository.

README

@altronis/tokenmark-mcp

An MCP server that gives Claude (and other agents) real local-LLM benchmark data + hardware-aware model recommendations from TokenMark.

Ask "what should I run on a Strix Halo for coding?" and the agent answers from measured configs (decode tok/s, quant, backend), each with a source link. It never invents numbers.

Add to Claude Code

sh
claude mcp add tokenmark -- npx -y @altronis/tokenmark-mcp

Add to any MCP client

Run the server over stdio:

sh
npx -y @altronis/tokenmark-mcp

Or in a client config:

json
{  "mcpServers": {    "tokenmark": {      "command": "npx",      "args": ["-y", "@altronis/tokenmark-mcp"]    }  }}

Tools

  • tokenmark_recommend: { hardware, tasks?, prefer?, limit? } → ranked model + best-config picks (5 by default, 10 max) with measured tok/s + why.
  • tokenmark_configs: { model?, hardware?, limit? } → tracked benchmark configs, fastest first (25 by default, 50 max; matched gives the full count), with the run mode (speculative: true for MTP/DFlash/draft-model runs).
  • tokenmark_hardware: { platform? } → the hardware catalogue (Strix Halo, Gorgon Halo, DGX Spark, Mac Max/Ultra). With a platform: chip specs with a source per value, the boxes that ship it, Singapore prices.
  • tokenmark_search: { term } → matching models/configs.
  • tokenmark_submit: { repo, source?, note? } → queues a GitHub/Hugging Face repo with benchmark numbers for human review.
  • tokenmark_submission_status: { id } → where a submission is.

A speed someone measured themselves, or hardware missing from the catalogue, goes through the signed-in form at https://tokenmark.app/submit.

Data is pulled live from https://tokenmark.app (override with TOKENMARK_URL). Zero runtime dependencies.

The same data in your terminal

The CLI lives in cli/ of this repo:

sh
npx @altronis/tokenmark-cli recommend --hardware "Strix Halo" --tasks coding

The files in bin/ are the published builds, made from the TokenMark tracker that runs https://tokenmark.app.

MIT licensed. Data aggregated from public community benchmarks with attribution.

Source: README.md at commit 19596dc

Tools

0
Tool metadata has not been indexed yet.

Version history

1
  1. v0.2.1LatestOct 5, 2026