name: ask-llms
description: Query multiple LLMs in parallel via OpenRouter and compare their responses. Use when the user wants diverse perspectives from different models on a question.

Ask LLMs

Query multiple LLMs simultaneously via OpenRouter and display their responses side-by-side. For quick parallel comparisons — use /llm-council when you want models to actually debate and synthesize.

When to Use

Use this skill when:
- Comparing how different LLMs answer the same question
- Need diverse perspectives on a problem or decision
- Want quick "second opinions" from multiple models
- Curious about model differences on a topic
- Quick feedback on outreach messages (see /outreach for style tips)

Prerequisites

OpenRouter API key (required):

export OPENROUTER_API_KEY=sk-or-v1-...

Instructions

Step 1: Get the Question

Ask the user what question they want answered, or use the question they already provided.

Step 2: Run the Query

Expensive mode (default) — frontier models with thinking:

cd /Users/terry/skills/ask-llms
uv run council.py "YOUR QUESTION HERE"

Cheap mode — fast and affordable:

uv run council.py "YOUR QUESTION HERE" --cheap

Step 3: Present Results

The script outputs each model's response with clear separators. Highlight:
- Key differences between models
- Points of agreement
- Which response is most relevant to the user's situation

Model Tiers

Expensive (default) — 2026 frontier models with thinking/reasoning:
- Claude Opus 4.5
- GPT-5.2
- Gemini 3 Pro
- Grok 4
- DeepSeek R1

Cheap (--cheap flag) — fast and affordable:
- Claude Sonnet 4.5
- GPT-4o
- Gemini Flash
- Grok 4.1 Fast
- DeepSeek V3.2

Options

vs /llm-council

Files

• Script: /Users/terry/skills/ask-llms/council.py
• This skill: /Users/terry/skills/ask-llms/SKILL.md