A free home for everyday AI tools

npmuzei is a free, no-signup directory of 300 AI tools organised across 21 categories — writing, business, education, marketing, fun, gaming, translators and more. Pick a tool, type what you need, and stream back a result in seconds.

What we are (and what we aren't)

We are a thin, friendly interface on top of open AI models served across a curated pool of AI model providers. Every prompt you write goes through one of the free large language models in the dropdown — Llama, Mistral, Gemini Flash, DeepSeek, Qwen, and others as they become available.

We aren't a model lab. We don't train models. We don't store your conversations. We don't sell your data. The site is a tool, not a product trying to upsell you.

How the daily free quota works

Every visitor gets 1,000,000 free tokens per day, shared across all 300+ tools. A token is roughly 0.75 words, so a daily quota is enough for about 750,000 words of generated text — comfortably more than any single person can read or write in a day.

We track usage through a combination of a signed browser cookie, your IP address, and your network subnet. The quota resets at 00:00 UTC every day. When the rotating pool of free API keys is busy, the request retries up to five times across different keys before returning a friendly error.

Privacy first

• No accounts. We don't ask for an email, phone, or anything that identifies you.
• No chat storage. Conversations live only inside your browser tab. Refresh the page and they're gone — we never write them to disk.
• No tracking pixels. No Google Analytics, no Facebook Pixel, no third-party ad networks.
• Stable cookie is used only to enforce the daily quota — nothing else is stored against it.

See the Privacy Policy for the full detail.

Who built this

npmuzei is an independent project. The codebase is a Next.js app with a SQLite-backed tool registry, multi-key API rotation, and per-IP/cookie/subnet token quotas. Articles for each of the 300 tools are hand-written, not auto-generated, so you get real guidance instead of fluff.

Why keep it free?

The cost of inference on free open models is genuinely small. By pooling requests across multiple provider accounts we can serve a meaningful daily quota to every visitor without charging. If usage ever grows beyond what the free tier supports, we'll add an optional paid tier — but the existing free tier will stay free.

Contact

Have a tool idea, found a bug, or want to report abuse? Head to the contact page.