The closest thing to a hands-free intelligent assistant on a phone.
Consilix rank 9/10
What it is
Native speech-in, speech-out conversation built on a multimodal model that does not transcribe to text in the middle. Sub-three-second latency, interruptible, with intonation and emotional nuance. Vision is integrated so you can point a camera at something while talking.
Key features
Sub-three-second response with natural interruption handling
Live camera input during voice conversation
Real-time translation between languages, replacing standalone apps
Voice style instructions, for example whisper, slower, more excited
Seamless switching between voice and text mid-thread
Best use cases
Walking a site with the camera on and asking about defects in real time
Live translating a conversation with a Polish or Romanian subcontractor
Practising a pitch out loud and getting coached on pacing
Hands-free dictation while driving or commuting
Reading a French menu or a German tax form through the camera
Weaknesses
Will not sing or imitate named voices. Background noise still trips it more than the demos suggest. UK regional accents (Glaswegian, strong Scouse, Belfast) sometimes get mistranscribed.
Pricing
Short daily preview on Free. Materially higher caps on Plus and Pro. Available in app and web.