Answer engines can only cite you if they can read you. Bloggable publishes your blog in the formats they look for, and gives you explicit control over who's allowed in.
What every blog serves
llms.txt — at your blog's root. A short, machine-readable menu of the blog and its best posts, in the emerging convention answer engines look for when asked "what is this site?".
A .md version of every post — the same article as clean markdown, at the post's URL with .md appended, and advertised to crawlers. HTML is full of navigation, scripts and layout; markdown is the article. Giving a model the markdown means it quotes your writing rather than your page furniture.
Both are on by default. The markdown variants can be switched off per blog if you'd rather not serve them.
Controlling which AI crawlers may read the blog
Per blog, in the SEO settings, set the AI-crawler policy:
- Allow — an explicit permission for eleven named AI crawlers: GPTBot, OAI-SearchBot, ChatGPT-User, ClaudeBot, anthropic-ai, PerplexityBot, Google-Extended, CCBot, Bytespider, Amazonbot and Applebot-Extended.
- Block — disallow them, and suppress
llms.txt. - Default — emit no named rules and let the wildcard robots.txt rule apply.
This is a real trade-off, not a formality. Blocking keeps your writing out of training data and out of AI answers. Allowing is how you get cited. Most people publishing in order to be found should allow; if the content is the product rather than the marketing for it, blocking may be right.
One honest limit: robots.txt is a request, not a fence. Well-behaved crawlers respect it.
For your readers
Posts can carry an "Ask AI about this article" link that opens the post in ChatGPT, Claude, Perplexity or Kagi. It's on by default, per blog.
Bloggable's own
We serve what we sell. Bloggable's llms.txt is at /llms.txt, every help article has a .md version, and the whole help centre is available as one file at /help/llms-full.txt — the Copy page button on any help article hands it to a model directly.