Mintlify Pricing: Docs Hosting and llms.txt

There is a specific way this purchase goes wrong. A team reads that AI assistants are starting to fetch llms.txt, finds that Mintlify ships it automatically, and starts pricing a migration of their entire documentation site around a file that is thirty lines of markdown. Six weeks later they have moved their docs onto a new platform to solve a problem that was never the expensive part.

So this is a pricing breakdown with the decision put back in the right order. Every number below was read from the Mintlify pricing page and its own llms.txt documentation in September 2026, not from an aggregator, because software directories in this category go stale within weeks of a repricing.

Mintlify pricing tiers

Three tiers. What changes between them is seats, AI capability and governance, in that order.

Annual billing is offered on Pro. The jump from $0 to $450 is the whole ladder for most companies, since Enterprise is a procurement conversation rather than a price. That is a wide gap, and there is nothing in between, which is the single most important fact about this pricing model: there is no $99 tier for a six person team that outgrew five seats.

What you get for llms.txt at $0

This is the part worth being precise about, because it is where the money either gets spent well or wasted.

Mintlify automatically hosts an llms.txt file at the root of your project listing all available pages, and an llms-full.txt alongside it. Both are also served at the /.well-known/ paths for tools that look there. Page links in the index carry a .md extension so an agent can fetch the markdown version of a page directly rather than parsing your rendered HTML. Descriptions in the index use the first paragraph of each page description, truncated at 300 characters. When the index would exceed 100,000 characters, Mintlify splits it under /_llms/ and may shorten descriptions further to keep each file under the limit.

None of that is gated behind a paid plan in the documentation. If your docs sit on Mintlify, the AI files are handled, at every tier, including the free one.

Which reframes the question. You are not buying llms.txt. You are buying a documentation platform, and the files come along.

What the $450 actually buys

Pro is an AI and team plan, not a file plan. The concrete additions are unlimited editor seats, the docs assistant and the agent, automations, preview deployments on pull requests, admin APIs and platform analytics, plus 10,000 AI credits a month with overages at a cent each.

Two of those tend to decide it. The first is seats: five editors is generous for a startup and tight for a company where support, solutions engineering and three product teams all edit docs. The second is preview deployments, because a docs change you can review in a pull request is a different workflow from one you find out about after it ships.

The AI credits are the part to model rather than assume. Ten thousand a month sounds like a lot until the assistant is answering questions for every visitor on a busy docs site. At a cent per credit over the allowance, an unmodelled month is not catastrophic, but it is not zero either, and it is the only variable line on the bill.

When Mintlify is the right buy

If your documentation is your product surface, it is written in MDX, and you want hosting, search, an API playground, an MCP server and the AI files handled without owning any of it, Mintlify is a good purchase and the free tier is unusually generous. A built-in feature beats an integration every time, and we would rather say that plainly than pretend otherwise.

It is the wrong buy in three situations, and they are common enough to name.

The first is when your docs are not the only content that matters. Plenty of teams have reference docs on one platform, a knowledge base on another and a marketing site on a third, and an llms.txt that describes one third of your public surface is a partial answer. The same split shows up in content operations generally, which is why teams that solved docs hosting often still have a marketing site where keeping a publishing schedule running without a full time writer is an unrelated and unsolved problem.

The second is when you want a markdown corpus of documentation you do not own. Building a support agent that answers questions about the three vendor APIs your product integrates with means ingesting their docs, and no platform decision of yours affects their site. That is a crawl.

The third is a migration you cannot justify. If your docs already run on Docusaurus, MkDocs, a Next.js site or a CMS, and everything works, moving platforms to obtain two generated files is the most expensive possible way to get them. Generating llms.txt and llms-full.txt from your existing site is a build step, not a replatform.

The cost of doing it without a docs platform

For comparison, here is what the same outcome costs when you keep the docs you have. The variable is how many pages you need converted to clean markdown and how often.

The third row is the one people forget to consider, and it is free. If your documentation lives in a repository as markdown already, you can build both files from source at publish time without crawling anything or paying anyone. Do that. It is the correct answer for a static site generator with markdown sources, and no vendor benefits from telling you so.

Crawling earns its place when the markdown is not sitting there ready. That is the case when the docs render client-side, when they are spread across several properties, when the source is a CMS rather than files, or when the pages belong to somebody else. Then you point an llms.txt generator built on a live crawl at the tree, take one clean markdown document per page, and assemble the index and the corpus from the result.

Is llms.txt worth paying anything for?

Be clear-eyed about the return, because the marketing around this file has run ahead of the evidence. Google Search Central states that Google Search does not require new machine readable files to appear in Search including its generative AI features, that Search ignores llms.txt, and that creating one will neither harm nor help your rankings. Google also says it is completely fine to maintain the file for other services that use it.

So the case for llms.txt is not organic search. It is that developer-facing tools and agent frameworks do fetch it, that Anthropic, Stripe, Cursor, Cloudflare, Vercel, Supabase and Mintlify itself publish one, and that a clean markdown corpus of your own documentation is useful to your own retrieval systems whether or not any third party ever reads the file.

Which is the honest frame for this pricing decision. Nobody should pay $450 a month for llms.txt. Plenty of teams should pay $450 a month for a docs platform with unlimited editors, preview deployments and an assistant, and take the files as a bonus.

How much does Mintlify cost per month?

Zero on Starter, $450 on Pro, and a quoted figure on Enterprise. Starter includes five editor seats, a custom domain, the web editor, authentication, an MCP server and the API playground, which is a real product rather than a trial. Pro adds unlimited seats, the assistant and agent, automations, preview deployments and 10,000 monthly AI credits with overages at $0.01 each.

Does Mintlify generate llms.txt automatically?

Yes, on every project. Mintlify hosts llms.txt at your project root listing all available pages, plus llms-full.txt with the page content inlined, and mirrors both at the /.well-known/ paths. Links in the index carry a .md extension so agents can fetch markdown directly. The documentation does not gate this behind a paid tier.

What happens when a docs site is too large for one llms.txt?

Mintlify splits it. When the index would exceed 100,000 characters, it publishes split index files under /_llms/ and may shorten page descriptions to keep each file under the character limit. That is worth knowing before you assume a single fetch gives an agent your whole map, because on a large site it does not, and any tooling you build against the file needs to follow the split.

Can I use Mintlify just for llms.txt without moving my docs?

No. The files are generated from the documentation hosted on the platform, so there is nothing to generate until your content is there. If moving is not on the table, generate the files from your own source at build time when you hold the markdown, or crawl your live site to markdown when you do not. Both leave your existing docs stack alone.

The decision, in one line

Price the documentation platform on seats, review workflow and the assistant, because that is what the $450 is for. Treat llms.txt and llms-full.txt as a checkbox that several routes satisfy, including two that cost nothing. If you are staying put, the files are a build step; if you are moving anyway, Mintlify hands them to you at the free tier.