tinyhumans.ai

mx

smtp.google.com

ns

clayton.ns.cloudflare.com

mona.ns.cloudflare.com

caa

txt

google-site-verification=laYmayMjsjR6tqPdZIXSRKgOhBlx_ES05Bd7OXNUMO4

facebook-domain-verification=2qqgqwy2wt5zz4411s0bgug8g7p0oa

tiktok-developers-site-verification=4Ji0nlsXxzuuXVugABt1bdFQbw5kxfxg

v=spf1 include:_spf.google.com include:amazonses.com ~all

tiktok-developers-site-verification=9NusYWikqcXMEnenhFvqkhjD772ON0XH

dmarc

v=DMARC1; p=quarantine; rua=mailto:security@tinyhumans.ai

User-Agent: *
Allow: /
Disallow: /api/

User-Agent: GPTBot
Allow: /
Disallow: /api/

User-Agent: OAI-SearchBot
Allow: /
Disallow: /api/

User-Agent: ChatGPT-User
Allow: /
Disallow: /api/

User-Agent: ClaudeBot
Allow: /
Disallow: /api/

User-Agent: Claude-SearchBot
Allow: /
Disallow: /api/

User-Agent: Claude-User
Allow: /
Disallow: /api/

User-Agent: anthropic-ai
Allow: /
Disallow: /api/

User-Agent: PerplexityBot
Allow: /
Disallow: /api/

User-Agent: Perplexity-User
Allow: /
Disallow: /api/

User-Agent: Google-Extended
Allow: /
Disallow: /api/

User-Agent: Applebot-Extended
Allow: /
Disallow: /api/

User-Agent: Amazonbot
Allow: /
Disallow: /api/

User-Agent: Meta-ExternalAgent
Allow: /
Disallow: /api/

User-Agent: CCBot
Allow: /
Disallow: /api/

User-Agent: cohere-ai
Allow: /
Disallow: /api/

User-Agent: Bytespider
Allow: /
Disallow: /api/

Sitemap: https://tinyhumans.ai/sitemap.xml

# TinyHumans

> TinyHumans builds open source AI agents you run yourself. Three products: OpenHuman, a local-first personal AI for Mac, Windows and Linux; OpenCompany, an agent hive mind that reasons as a team; and Medulla, a terminal app that runs every coding agent you have at once.

Everything we ship is open source and designed to run on your own machines with your own models. Data stays local by default; see the privacy policy for what the hosted services do and do not retain.

## Products

- [OpenHuman](https://tinyhumans.ai/openhuman): GPL-3.0 personal AI for macOS, Windows and Linux. Fast, cheap and intelligent, in that order: a Rust core running in-process (102ms cold turn, ~25x denser than one process per agent), one key covering inference, search, embeddings and integrations instead of five vendor accounts, and local-first memory that recalls with citations before every turn. 119 managed-auth OAuth toolkits, 26 BYOK providers plus Ollama/LM Studio/MLX. Early beta.
- [OpenCompany](https://tinyhumans.ai/opencompany): GPL-3.0 agent hive mind. A team of AI specialists that give independent first takes, then options backed by evidence, then a decision you can follow line by line. Self-hosted or hosted.
- [Medulla](https://tinyhumans.ai/medulla): GPL-3.0 terminal that runs Claude Code, Codex, OpenCode and plain shells side by side, locally or on any machine you can SSH to. Reads each session for the moments needing a person — permission prompts, usage limits, crashes, finished turns — and surfaces them as one count.

## Pricing

- [Pricing](https://tinyhumans.ai/pricing): Start free with local models. Paid tiers add Medulla orchestration, seats and credits; a flat monthly price buys slightly more than its value in credits, metered per API call.

## Writing

- [Why agents should answer before they read the room](https://tinyhumans.ai/blog/why-agents-should-answer-before-they-read-the-room): Three agents agreeing is usually one guess repeated. A blind first round and one change of wording took a team of agents from 16% to 67% correct.

## Source

- [GitHub](https://github.com/tinyhumansai): every product's source, issues and releases.

## Legal

- [Privacy policy](https://tinyhumans.ai/privacy): what we collect, how long we keep it, and your rights.
- [Terms](https://tinyhumans.ai/terms): terms of service for OpenHuman and related services.