Open-source voice AI platform you can self-host — a visual workflow builder, bring-your-own STT/LLM/TTS models, telephony, and no per-minute platform lock-in.
Freemium
Summary
Developer platform for AI agents on Apple's own channels — programmable iMessage at scale, plus video agents that answer and place FaceTime calls in a few lines of code.
Screenshots
Description
Chert makes Apple's messaging stack programmable. The core product is an API and workflow layer for sending, receiving, routing and observing real iMessage conversations at scale; the newer FaceTime product extends the same idea to video, letting an agent answer or place a FaceTime call.
Why video matters here
The pitch for the FaceTime product is blunt: "Voice agents are blind. Chert isn't." A phone agent troubleshooting hardware has to work from a caller's description — the site's own example is a user saying "Uh, which cable?" On a video call the agent sees the cable. That turns a class of support, onboarding and field-service conversations from a guessing game into something the model can actually resolve.
Building an agent
Configuration is three steps: write the instructions and pick the realtime model, voice and behavioural parameters; choose the avatar and framing for the persona; then publish and assign the assistant to a provisioned FaceTime line. A browser preview lets you exercise prompts, voice, microphone behaviour, avatar and settings before anything is assigned to a real line, and both inbound and outbound calls can be tested.
Use cases
On the iMessage side: customer onboarding, support ticketing, conversational AI agents and outbound sales. On the FaceTime side: anything where showing beats describing.
Availability
The FaceTime product is in controlled private preview while Chert validates production readiness — acceptance mode is off by default and live execution requires explicit authorisation. No public pricing is listed; access is arranged with the company. Chert is Y Combinator-backed, with investment from Z Fellows and Betafund and angels from NVIDIA, LinkedIn, OpenAI, Anthropic, Scale AI and Grafana.
Worth knowing
Operating on iMessage and FaceTime means operating inside Apple's rules and a user expectation that these channels are personal. Volume outreach here carries reputational and compliance weight that email does not.
Reviews
Similar App Suggestions
Voice-first personal AI agents with continuous memory across voice and text — four distinct personalities, ultra-low latency, free on iOS.
Free
Local-first meeting and interview recorder that transcribes, labels speakers and writes up your notes entirely on your own computer, for a one-time price.
Paid
Local-first Mac dictation that types into any app, plus file and link transcription across 110 spoken languages on Apple Silicon.
Freemium
Voice dictation that works in every app and cleans up as you speak — no filler words, correct formatting, right tone.
Freemium