Infini-AI-Lab

Batch Benchmark — Development skill for Claude Code

Development community

Launch the 4-variant batch on the currently-free local GPUs (parallel when N>=4, otherwise sequential waves of N).

How to install Batch Benchmark

Installs to ~/.claude/commands/infini-ai-lab-vortex_torch-batch-benchmark.md

Terminal

mkdir -p ~/.claude/commands && curl -fsSL https://raw.githubusercontent.com/Infini-AI-Lab/vortex_torch/HEAD/.claude/commands/batch-benchmark.md -o ~/.claude/commands/infini-ai-lab-vortex_torch-batch-benchmark.md

Restart Claude Code, or start a new session, for it to be picked up.

What Batch Benchmark does

description: Launch the 4-variant batch on the currently-free local GPUs (parallel when N>=4, otherwise sequential waves of N). The only sanctioned benchmark form. Pass exactly 4 submission names. argument-hint: (always 4 variants; parallelism = min(N_free_gpus, 4))

Run the **4-variant batch** against the AIME24 benchmark. **Batch size is fixed at 4.** Parallelism depends on free-GPU count `N`:

N >= 4 → all 4 variants run in parallel, one per GPU.

Alternatives in Development

Bypass 403 — Probe a 403/401 endpoint with the most-paid bypass tricks (header injection, path encoding, method swap, WAF f 4.5k ★
/report — Generate a submission-ready bug bounty report 1.8k ★
Axiom Marketplace Submission Guide — Axiom Marketplace Submission Guide skill 834 ★