Checks your hardware first
It reads your processor, memory and free disk, then shows which models will run well.
NevTan AIFree
Open AI models, running on your own computer
NevTan AI checks what your hardware can handle, installs the model for you and answers locally. No subscription, and nothing charged per message.
- Answers stay on your device
- Free, no per-message cost
- Available on macOS
Local AI without the setup
Running a model normally means file formats, memory limits and a command line. NevTan AI does that for you.
One-click install
Pick a model for the task in front of you and install it in one click.
Chat and coding models
Qwen, Llama, Gemma, Mistral and Phi for chat, and Qwen Coder, DeepSeek Coder, Devstral and CodeGemma for code.
Conversations you can find
Conversations are saved, searchable and easy to pick up again.
Switch models mid-chat
Move a conversation to a bigger model without starting over.
Use it from your editor
Create an API key and point any OpenAI-compatible tool at NevTan AI.
A closer look at NevTan AI
Made for your machine
- Reads processor, memory and free disk
- Recommends models that will run well
- Shows requirements before you download
- Keeps track of model files and updates
Security
Your machine, your rules
NevTan AI runs on hardware you own and asks before it acts.
Local answers
Local models generate answers on your own computer.
Asks first
NevTan AI asks before it changes anything on your computer.
Keys shown once
Copy a new API key when you create it. It can’t be shown again.
Revocable keys
Anyone with a key can use your model quota, so revoke keys you no longer use.
Short-lived pairing codes
A device pairing code works once and expires in ten minutes.
One NevTan account
The same sign-in you use for every NevTan product.
NevTan AI questions
How much does NevTan AI cost?
Nothing. There is no subscription and nothing is charged per message for models it runs locally. Hosted models are billed by whichever provider issued the API key you connect.
Which computers does it run on?
NevTan AI is available on macOS today. Windows and Linux downloads are being built now.
Does it use my graphics card?
It detects your graphics card today, but GPU acceleration is still being built. Recommendations are based on system memory for now.
Can I use a hosted model instead?
Yes. Connect a hosted model with your own API key, and that provider bills you directly.
Is NevTan AI part of my NevTan account?
Yes. One account works across every NevTan product.
Get started with NevTan AI
Free for everyone with a NevTan account, on every plan.