Blog

Thoughts on voice, AI, and the craft of making tools that disappear — from a small team of algorithm researchers and daily AI-product users.

Omni-modal voice typing: multimodal understanding, MoE, and streaming text output

Omni-modal voice typing needs a split pipeline, streaming audio, and Mac-local context. A look inside Loqua's voice stack.

Compare

How-to

Engineering

Productivity