mythos.one

What it is

MythOS is an AI-Native Knowledge Platform. At its core, it is not a conventional note-taking app, but a “shared memory layer” that turns personal memos, tags, mentions, public content, and AI conversations into knowledge that can be used by multiple models. It is aimed at writers, researchers, founders, and people who think in public, with an emphasis on ensuring your knowledge base is not locked to Claude, ChatGPT, or any single model.

Core features

At the “Memory” layer, MythOS provides a Markdown editor with support for @mentions, #hashtags, collapsible headings, image pasting, and automatic backlinking, making it suitable for long-term personal knowledge management. At the “Mind” layer, it supports semantic search and conversations with your personal memos; answers cite the specific memo sources, reducing the limitations of pure keyword search. At the “Mouth” layer, memos can be published with one click as public pages, newsletters, or community discussions, and are exposed to search engines and AI indexers via JSON-LD and /llms.txt.

AI, API, and integrations

MythOS stands out for being model-agnostic and agent-friendly. Its materials indicate support for MCP server, Agent API, OpenAPI, and /llms.txt. It can connect to Claude Desktop, Claude Code, Cursor, and Windsurf, and can also be read by Zapier, n8n, or custom scripts. Its AI capabilities rely on third-party services such as OpenAI and Anthropic, support bring-your-own API keys, and store keys in encrypted form.

Pricing and privacy

Pricing is split into three tiers: Scribe costs $10/month and is suitable for personal note-taking; Scholar costs $50/month and unlocks key features such as public publishing, collaborators, Agent API, MCP, and email import; Oracle is a custom plan for communities, custom domains, and custom AI Agents. On privacy, memos are private by default, users retain ownership of their content, and Markdown and media can be exported. The company states that it does not train on user libraries or sell context. However, after account deletion, some public or community content may be anonymized and retained, so it should be deleted separately in advance if needed.

Pros, cons, and who it’s for

Its strengths are portability, open interfaces, and suitability for long-term context management across models. The downsides are that no free trial is disclosed, meaningful AI/Agent integrations start from the Scholar tier, the entry cost is not low, and AI output is not guaranteed to be accurate. It is best suited to heavy writing, research, knowledge publishing, and automated workflow users. For lightweight note-taking, Notion, Obsidian, or Logseq may be simpler.

Access from China

The collected text does not specify access from mainland China, payment methods, or localization support, so this is currently unknown. If the service depends on third-party capabilities such as OpenAI, Anthropic, or Claude, users in China may also need to consider network access, account availability, and payment availability.