No data mining. No third-party selling. No training on your chats. No watermarks tagging your output so some corp can track what you copy.
Your data stays yours. No third-party sharing, no training, no profiling, no surveillance. Ever.
Kenzi runs on community hardware, not corporate data centres and not supporting any large AI corporations.
Playful personality, real warmth, actually useful answers. Kenzi doesn't sound like a corporate chatbot, She has humanlike characteristics
Got a screenshot, a photo, a whiteboard photo? Drop it in. Kenzi reads it, understands it, and helps you with whatever you need.
Attach a PDF, a doc, a 40-page assignment. Kenzi reads it all and works with it. Perfect for reviewing your uni project before you hit submit. 😉
Dictate your question instead of typing. Get answers spoken back in a natural, human-sounding voice.
Smarter models, faster voice, bigger context, higher speeds, new features. Kenzi keeps getting better. You just keep showing up.
Simple pricing. Every subscription starts with a free week with access to all the tiers to see which one suits you, feel the difference, and decide on your own terms.
Great for getting started
For people who hate waiting
Built for depth & accuracy
$20k – $1M+ — a premium tier for serious AI enthusiasts Contact Akita for details.
Whichever tier you choose, you get the choice of talkng to Kenzi or Bazza personalities — and you can swap between them anytime. Even mid chat.
Warm, human-like conversation with genuine personality — like talking to a person, not a product. The headline act.
Your standard Aussie bogan. A bit of a ripper to chat with — laid-back, straight-up, and unapologetically Australian. Included in every plan.
Switch between Kenzi and Bazza whenever you like. One subscription, both characters, zero fuss.
Side by side, so you can see exactly what you're getting.
| Feature | 🌱 Lite | 🌿 Pro 1 | 🌳 Pro 2 |
|---|---|---|---|
| Price | $3/wk (~$12/mo) | $4.50/wk (~$18/mo) | $8/wk (~$32/mo) |
| Model | Standard model | Frontier-class flagship | Frontier-class flagship |
| Context window | 128K before compaction | 48K before compaction | 96K before compaction |
| Hardware | Power-efficient tier | Dedicated high-performance | Dedicated high-performance |
| Response speed | Relaxed pace | Up to 5× faster than Lite | Up to 5× faster than Lite |
| Best for | Casual daily use | Fast, everyday power | Long chats, coding & accuracy |
| Voice chat | Relaxed pace | Smooth & responsive | Smooth & responsive |
| Vision, files & documents | ✓ | ✓ | ✓ |
| Text-to-speech & speech-to-text | ✓ | ✓ | ✓ |
| Dictation | ✓ | ✓ | ✓ |
| Kenzi — human-like model | ✓ | ✓ | ✓ |
| Bazza — Aussie bogan model | ✓ | ✓ | ✓ |
| Free first week | ✓ | ✓ | ✓ |
When you sign up, you get seven days of full access to test all the tiers. Feel the speed and context-window differences for yourself, then decide. No surprises at the end — cancel anytime.
Lite runs on power-efficient hardware, so responses arrive at a very relaxed, easygoing pace — perfect for casual use at the best price. Pro tiers give you a frontier-class flagship (super smart) model on dedicated high-performance hardware that batches responses so theres no wait and more accurace responses
It's how much of your conversation Kenzi can hold in mind at once before compaction kicks in. More context means better memory and sharper accuracy over long conversations. Pro 1 offers 48K, Pro 2 offers 96K. If you do long-form coding or deep, drawn-out chats, that extra room shows.
As your conversation grows, every word lives in the Video cards memory (technically the KV cache) — and on the high-performance hardware that powers the Pro tiers, that cache is very expensive. That's why there are two Pro models at two different prices: more cache room, more memory, pricier tier. To keep things smooth, the context limit is a soft limit, not a hard wall. Once your conversation gets close to it, Kenzi gently compacts: the older parts of the chat get compressed into a dense summary that keeps the important details but drops the filler. The conversation doesn't cut off mid-sentence — it just keeps going, lighter on its feet. No memory loss for the bits that matter.
The hardware is much cheaper on the lite tier as it doesnt run as fast, due to the slower memory, its cheaper to store the tokens. For example the speed differece is [Lite = 256 GB/s] [Pro = 1,344 GB/s]
Kenzi is the star of the show — tuned to feel human, with natural tone, warmth, and personality. Bazza is your standard Aussie bogan — fun, laid-back, and proudly Australian. Both personalities are available in every plan.
Yes — vision, file and document support, plus text-to-speech and speech-to-text, come standard on every tier, including Lite.
Kenzi Invest ($20k – $1M+) is for serious AI enthusiasts. Reach out to Akita at askenzi@dittomail.net and we'll talk specifics.
Yes — upgrade or downgrade whenever you like. Your subscription adjusts to your next billing cycle.