Private, offline LLM inference on iOS. No cloud. No subscriptions. One-time purchase, deploy forever.
Solo iOS developers building on-device AI.
Engineering teams shipping production iOS apps.
Large organizations with custom requirements.
Buy once, deploy forever. All licenses are perpetual. You own the source code. No vendor lock-in. No recurring charges. Security updates and minor versions are always free.
iOS is fully supported (including iPad). macOS support is in development. watchOS coming Q4 2026. Enterprise customers can request early access.
Source code is included. Fork it, modify it, ship it. No restrictions. For large-scale customization, enterprise plans include dedicated engineering time.
Developer: 10 apps. Team: 100 apps. Enterprise: unlimited. These are soft limits. Contact us if you need flexibility.
Yes. Developer and team licenses are one-time purchases. Enterprise can be structured as annual subscriptions with volume discounts and service agreements.
Any GGML-quantized model: Qwen, Llama 2/3, Mistral, Phi, and more. We pre-optimize for popular models. Custom quantization available for enterprise.
30-day money-back guarantee on all tiers. After 30 days, licenses are perpetual and non-refundable. You keep the SDK forever.