ALL open-source text models are included! Popular examples:

Note: TEE variants of models not included yet. We guarantee 99.9% availability for popular models like DeepSeek V3 and 3.1. Some roleplaying finetuned models may rely on a single provider, which can occasionally affect availability.

What does "unlimited personal access" mean?

Practically speaking, we cap at 60,000 generations per month, or 2,000 per day (you can choose for yourself). You can run requests non-stop based on 1 request every 30 seconds that you're awake, every day of the month - if you think you would use more than this for personal use let us know and we will reconsider whether this limit should be higher. What we are mostly trying to discourage is people using this subscription for commercial purposes.

Can I use this via API and website?

Yes! Your subscription works seamlessly across both our website and API. Use the same quota whether you're chatting on the web or integrating with your applications.

Is my data private?

Absolutely. None of your data is stored or logged on our servers - everything is saved locally on your device only. Additionally, none of the providers we use for open-source models log or store prompts, nor use them as training data. Your conversations remain completely private.

Can I still use paid models with a subscription?

Yes! The subscription gives you free access to open-source models, but you can still use premium models like GPT-5, Claude, Gemini and others with your regular balance.

What happens if I exceed my quota?

You can choose to exceed your daily 2,000 query limit - the only hard cap is 60,000 queries per month. We designed it this way for flexibility, since some days you might use AI heavily and other days less. Once you hit the 60,000 monthly limit, you'll need to wait for the monthly reset or use your regular balance for continued access.

Does this include video generation?

No, video generation models are not included in the subscription. They require separate payment using your balance.