Onde

On-device inference for cross-platform apps.
Run LLMs, diffusion models, and speech-to-text locally — no cloud, no latency, no data leaving the device.

Onde is a Rust crate that gives your app on-device AI in a single dependency. It wraps mistral.rs for LLM and image generation, and whisper.cpp (via transcribe-rs) for speech-to-text — with a unified API that handles model discovery, HuggingFace Hub downloads, cache management, and GPU acceleration across every platform.

#License

MIT License

#Copyright

2026 Onde Inference