Benchmark Model
modular/skills/benchmark-modelby modularb9b3a8e86700No license200 starsListed Oct 8, 2026Updated Oct 8, 2026Repository updated today
Benchmark a model served on MAX with the `max benchmark` command: measure throughput (tokens/sec), latency (TTFT, TPOT, inter-token latency), and GPU utilization by driving load against a running `max serve` endpoint. Use this whenever the user wants to benchmark, load-test, or measure the performance of a MAX model, get tokens-per-second / TTFT / TPOT numbers, run a concurrency or request-rate sweep, compare latency vs throughput, size a deployment, or produce benchmark JSON, even if they don't say "benchmark" by name. Also use when a `max benchmark` run fails to connect or reports zero/garbage numbers.
Add to a SourceWeft workspace
- Open the skill in your dashboard and add it to a workspace.
- Enable it for the chats that should use it.
This skill is instructions only: it ships no scripts to execute.
Add to SourceWeftYou will be asked to sign in first, then taken straight to this skill.
Ask your agent to install it
Paste this prompt into Claude Code, Codex, Cursor or another agent that can run commands — or into SourceWeft chat. The agent reads this skill's install guide, shows you its source, license and scripts, and installs it with the SourceWeft CLI once you agree.
Read https://sourceweft.com/skills/gh-modular-skills-benchmark-model/install.md and install the skill it describes. Before installing, show me its source, license and whether it ships scripts, and wait for my OK. Ask me before changing anything else on my machine.Install it yourself from a terminal
For Claude Code, Codex, Cursor and other local agents. The SourceWeft CLI fetches the skill from its source repository at the commit scanned here, and verifies every file against the hashes recorded when the skill was scanned. If anything differs, nothing is written.
npx @sourceweft/cli skills install @modular/benchmark-modelAdd --agent claude-code, codex, cursor or universal to choose which agent gets it (Claude Code by default).
Upstream installer — not verified by SourceWeft
The open-source skills installer fetches the same pinned commit, but does not check the files against the hashes SourceWeft recorded.
npx skills add https://github.com/modular/skills/tree/b9b3a8e8670029b90b4842b8ab5a7d5f494108af/benchmark-modelSource and attribution
Source:modular/skillsinbenchmark-modelat commitb9b3a8e
License: No license
Content belongs to its original authors. SourceWeft indexes it from a public repository.
More from modular/skills
Mojo Syntax
modular
Guides writing and reviewing Mojo code with current syntax, replacing obsolete pretrained patterns.
New Modular Project
modular
Guides creation of a new Mojo or MAX project, choosing an environment manager and channel.
Mojo Gpu Fundamentals
modular
Teaches how to write Mojo GPU kernels, correcting CUDA-based assumptions about syntax, memory and launches.
More in DevOps & Cloud
Playwright Devops
microsoft
DevOps workflows for Playwright: analyze GitHub Actions failures for the last commit on main and fetch failed job logs.
M5 Onboard
anthropics
Provisions M5Stack ESP32 boards by detecting them on USB, flashing UIFlow 2.0 firmware, and installing a MicroPython app bundle.
Yeet
openai
Stages, commits, pushes, and opens or updates a GitHub pull request in one flow using the GitHub CLI.
Runbook
anthropics
Creates or updates step-by-step operational runbooks for recurring tasks, including troubleshooting, rollback and escalation.
Incident Response
anthropics
Guides an incident response workflow: severity triage, status updates, mitigation tracking, and blameless postmortems.
Deploy Checklist
anthropics
Generates a pre-deployment readiness checklist covering pre-deploy, deploy, post-deploy and rollback triggers.