Agent SkillsAtlasBETASign in
Back to directory

Inference Perf Bench

Agent skill
1 skill
READMEOverview ↓

Canonical inference perf-bench skill (formal name. The colloquial alias is ai-bench - identical behaviour). Drives NVIDIA AIPerf + the replay-playback dataset against an in-cluster vLLM endpoint to measure TTFT, ITL, throughput, tok/s/user, request latency, and prefix cache hit rate. Iterative 9-phase workflow. Use when promoting a model to staging/prod, after vllmArgs / vLLM / KV-cache-dtype changes, or for A/B comparison across configs. Triggers on "perf-bench", "AIPerf", "Replay Playback", "throughput sweep", "TTFT P95", "concurrency sweep on inference", "/run-perf-bench", "benchmark throughput", "benchmark latency", "run aiperf", "performance test", or any combination of "perf / latency / throughput / tps / TTFT / ITL" with "inference / vllm / serverless / bench".

Install with
Claude CodeCodexCursorGitHub CopilotGemini CLI

Supported by the Atlas install prompt. Skill-specific requirements vary.

Skill stats

Skills1CategoryGeneralStars0

About this skill

Canonical inference perf-bench skill (formal name. The colloquial alias is ai-bench - identical behaviour). Drives NVIDIA AIPerf + the replay-playback dataset against an in-cluster vLLM endpoint to measure TTFT, ITL, throughput, tok/s/user, request latency, and prefix cache hit rate. Iterative 9-phase workflow. Use when promoting a model to staging/prod, after vllmArgs / vLLM / KV-cache-dtype changes, or for A/B comparison across configs. Triggers on "perf-bench", "AIPerf", "Replay Playback", "throughput sweep", "TTFT P95", "concurrency sweep on inference", "/run-perf-bench", "benchmark throughput", "benchmark latency", "run aiperf", "performance test", or any combination of "perf / latency / throughput / tps / TTFT / ITL" with "inference / vllm / serverless / bench".

Quick start

  1. Copy the install prompt. Use the green button above this README.
  2. Send it to your agent. Paste it into your coding agent’s chat and send.
  3. Review the setup. Your agent checks the skill’s instructions, requirements, and installation scope.

Installation guide →

Review, license & listing details
Review the source before using this skill

This is an imported listing. We haven’t certified its safety, license, or compatibility. Skills may include executable scripts.

Automated security scan: Partial

Only SKILL.md was scanned. Supporting files and scripts were not checked. No safety grade has been assigned.

0 potential findings · Scanned 11/4/2026 · Scanner 0.2.1

Content SHA-256: 75d58613fdf2644645557b959ee339f99104a915b8169c83933cdd0e5b891d4f

About scan coverage and limitations
Original listing metadata

Original name

inference-perf-bench

Original description

Canonical inference perf-bench skill (formal name. The colloquial alias is `ai-bench` - identical behaviour). Drives NVIDIA AIPerf + the replay-playback dataset against an in-cluster vLLM endpoint to measure TTFT, ITL, throughput, tok/s/user, request latency, and prefix cache hit rate. Iterative 9-phase workflow. Use when promoting a model to staging/prod, after vllmArgs / vLLM / KV-cache-dtype changes, or for A/B comparison across configs. Triggers on "perf-bench", "AIPerf", "Replay Playback", "throughput sweep", "TTFT P95", "concurrency sweep on inference", "/run-perf-bench", "benchmark throughput", "benchmark latency", "run aiperf", "performance test", or any combination of "perf / latency / throughput / tps / TTFT / ITL" with "inference / vllm / serverless / bench".

About this listing

Category
General
Reported license
MIT
Listing language (estimated)
English
Review status
Not individually reviewed
Agent compatibility
Atlas installation supports Claude Code, Codex, and Cursor. Skill-specific dependencies and features are checked during setup.
KEEP EXPLORING

Related packs & skills

Browse category →

Finding related packs and skills…

Inference Perf Bench | Agent Skills Atlas