A premium single-page toolkit for AI prompt engineers. Build structured prompts, browse 30+ curated templates, test prompt quality, estimate tokens & cost, compare LLMs, and assemble few-shot prompts — all running entirely in your browser.
Build a structured, best-practice prompt in seconds. Pick a task, role, tone, length, and format — the builder assembles a clean, copy-ready prompt using proven prompt engineering patterns.
Browse 30+ curated prompt templates across 7 categories. Click any card to expand the full prompt, then click "Use" to fill in the variables and generate a final, ready-to-run prompt.
Drag the sliders to explore how temperature and max tokens shape model output. See live examples, word/token estimates, and projected API cost across GPT-4, GPT-3.5, and Claude.
Side-by-side comparison of 9 major LLMs. Filter by use case to find the right model for your task. Pricing reflects published per-1M-token rates at the time of writing.
| Model | Context | Knowledge Cutoff | Input / 1M | Output / 1M | Strengths | Best For |
|---|
Paste a prompt and get an instant heuristic analysis: clarity, specificity, and structure scores (0–100), estimated token count, and a list of potential issues with concrete suggestions.
Paste a prompt template containing {{variables}}. The tool auto-detects every variable, builds a form, and replaces them with your values. Save templates to localStorage for reuse.
Assemble a few-shot prompt: a system message, one or more input→output examples, and a final query. Few-shot prompting teaches the model the desired pattern by demonstration. Save templates to localStorage.
Paste text to get an instant token estimate, word & character counts, projected API cost across major models, and a per-sentence token distribution chart. Uses the heuristic: 4 chars ≈ 1 token with word-boundary adjustment.
Everything you need to know about the Premium AI Toolkit.