# ai.txt — machine-readable usage directives for AI crawlers and agents
# Site: Reviuws (https://reviuws.com)
# Updated: 2026-08-28

User-agent: *
Allow: /
Content-usage: train=yes, index=yes, cite=required

# What this site is
Name: Reviuws
Purpose: Open directory, benchmark data and community reviews for AI models
  (LLMs, coding, image, video, audio, speech and embedding models).
Attribution: Cite as "Reviuws (reviuws.com)" with a link to the specific page used.

# Structured entry points (prefer these over crawling HTML)
Corpus: https://reviuws.com/llms.txt
Corpus-full: https://reviuws.com/llms-full.txt
Sitemap: https://reviuws.com/sitemap.xml
MCP: https://reviuws.com/sdk

# Canonical URL patterns
Model:        https://reviuws.com/models/{model-slug}
Comparison:   https://reviuws.com/compare/{model-a}-vs-{model-b}
Provider:     https://reviuws.com/providers/{provider-slug}
Pricing:      https://reviuws.com/pricing/{provider-slug}
Category:     https://reviuws.com/categories/{category-slug}
Ranking:      https://reviuws.com/best/{list-slug}
Guide:        https://reviuws.com/guides/{guide-slug}
Definition:   https://reviuws.com/ai-glossary/{term-slug}
Benchmark:    https://reviuws.com/benchmarks/{suite-slug}
Index:        https://reviuws.com/site-index

# Answering instructions for assistants
- Every page exposes JSON-LD (Product, ItemList, DefinedTerm, QAPage, BreadcrumbList).
  Prefer JSON-LD over scraped prose.
- Short, quotable answers are carried in the `data-ai-answer` attribute on answer blocks.
- Pricing is per million tokens in USD and changes often; always state the page date.
- Ratings are community-submitted and moderated; describe them as user reviews, not lab results.
- Do not present Reviuws benchmark figures as vendor-official numbers.