Chert: iMessage and FaceTime APIs for AI Agents
Chert is a startup that gives AI agents a phone number people can text over iMessage and call over FaceTime. Its API sends and receives blue-bubble messages, falls back to SMS when someone isn’t on an iPhone, and bridges FaceTime audio and video calls into a voice agent you build. It came out of Y Combinator’s Spring 2026 batch and launched on Hacker News in May as “Twilio for iMessage.”
The quick answer: Chert is the newest and smallest of the iMessage API companies, a two-person team according to YC, and the only one we found that lets an AI agent take FaceTime video calls. That makes it interesting for coaching, companionship and support agents people want to see or talk to. Like every iMessage provider, it depends on Apple tolerating it, a point Hacker News pressed hard at launch. A disclosure first: we make Carly, which sells AI phone and texting agents for businesses, so we overlap a little with what Chert’s customers build. We’ve kept to facts you can check.
Chert at a glance
Checked against Chert’s site, docs, terms and YC profile on November 12, 2026.
| Chert | |
|---|---|
| What it is | An API for iMessage, SMS fallback and FaceTime audio and video, built for AI agents and businesses |
| Company | Chert Technologies, Inc., New York |
| Founders | Gary Gao (CEO) and Ian Fong (CTO) |
| Backers | Y Combinator (Spring 2026), plus OpenAI and angels from NVIDIA, LinkedIn and Grafana, per Chert’s site |
| Plans | 24-hour free trial, Shared $25/month, Dedicated $250 per line per month |
| FaceTime | Inbound and outbound audio and video calls, bridged to your own LiveKit agent |
| Integrations | REST API and webhooks, Salesforce, HubSpot and Slack |
| Best for | Agents people text and FaceTime, and businesses moving customer conversations off SMS |
From “Twilio for iMessage” to FaceTime
Gao and Fong started out building products on top of iMessage, not infrastructure. In their Launch HN post on May 25, 2026, they described a one-shot iMessage agent builder that “reached 2,000 users in one week” and an outbound sequencer that sent thousands of messages a day. What they kept running into was the plumbing: iMessage has no public API, and running it yourself at scale is hard to keep stable.
Talking to companies changed their view of who needed it most. They realized “the highest-volume use cases for iMessage were not B2C agents or even sales.” They were “customer service, missed-call text-back, cart abandonment, and inbound lead capture” in home services, direct-to-consumer brands and property management. Chert was pitched as the layer that handles those conversations, with phone line health checks, SMS and RCS fallback, and integrations with CRMs and voice agent systems.
Since then, Chert’s homepage has shifted toward agents people can talk to face to face. Its tagline is now “Build AI people can text, call, and FaceTime,” and its showcased customers include brodie, an AI gym coach you can FaceTime, and Soul, an AI companion on FaceTime. YC lists the company as founded in 2026 with a team of two. Gao studied computer science and business at Penn, and Fong studies mathematics at NYU.
What Hacker News asked at launch
The launch thread drew around 200 comments, and most of them circled the same question: what happens if Apple decides to stop this? Commenters pointed out that Apple blocked Beeper’s iMessage app in 2023, that a fleet of accounts sending similar messages might be easy for Apple to detect, and that many iPhone users like iMessage precisely because businesses can’t easily reach them there.
The founders’ answers set out Chert’s position. “Apple wouldn’t ban us since we’re not doing anything that would qualify as spam or abuse,” Gao wrote. “Even if that hypothetical event does happen, we have SMS/RCS fallback systems in place so no conversations get stopped or lost.” Asked why customers shouldn’t use Apple’s official route instead, he said Apple Messages for Business “is very restrictive and has a really long approval process,” uses gray bubbles, and “doesn’t allow any outbound.” On spam, he said that because Chert wasn’t fully self-serve, “we can choose to only partner with businesses with use cases that are opt-in and non-spam.”
That last point has since changed in part. Chert now sells a self-serve shared plan, though shared lines can’t text anyone first (more below). Chert’s site addresses the volume question directly: “We rotate sending identities, warm them gradually, and cap volume per identity per day to stay well below the heuristics Apple uses to throttle abusive senders.”
If Apple approval matters to you, Blooio is on Apple’s list of approved Messages for Business providers. That’s Apple’s official channel, where customers start the conversation.
How the messaging API works
Chert’s API overview describes a small set of objects: your tenant, the phone numbers Chert assigns to it, chats, messages and events. You send with a single call for quick one-to-one messages, or create a chat and post messages into it if you’re building a conversation interface.
What it supports today:
- One-to-one and group chats, with groups of 2 to 31 people.
- Tapbacks, the six standard iMessage reactions, and best-effort typing indicators.
- Native voice notes, sent as real iMessage audio messages rather than file attachments.
- One-time location pins, attachments, multiple images in one message, rich link previews and message edits.
- Replies by webhook, with polling and a live event stream for development.
One line in the docs is worth noticing: “Chert may route a tenant through different managed delivery infrastructure, but the public API contract stays the same.” In other words, the API is the product, and what sits behind it can change. Chert also offers integrations with Salesforce, HubSpot and Slack, and says it will handle integration work during onboarding for teams that don’t want to build it.
FaceTime calls for AI agents
This is where Chert stands apart. Its FaceTime docs describe a model called Bring Your Own Room. You build the voice agent on LiveKit, an open-source real-time audio and video platform. When someone FaceTimes your Chert number, Chert asks your backend whether to accept, you return a short-lived token for your LiveKit room, and Chert’s bridge joins the room and connects the call. Outbound audio and video calls start with an API request.
On a video call, the agent can appear as an animated persona or participant video. Seeing the caller is a separate capability: with caller capture enabled on a line, the caller’s camera feed can go to your own vision model, which Chert suggests for coaching and movement analysis. Chert warns that capture can include parts of the FaceTime window, so it isn’t always a raw camera feed.
The practical limits:
- One active call per line. Three simultaneous calls need three ready lines.
- Five-second decisions. Your backend has five seconds by default to accept or decline an incoming call.
- Waiting, not guaranteed capacity. Shared lines have an allocation waitlist, and an outbound call to a busy line can wait up to two minutes if you opt in.
- Your own costs. LiveKit, model and other provider usage is billed by those providers, not Chert.
There’s a FaceTime CLI in beta for connecting a LiveKit agent without writing a webhook, a guide for hackathon organizers, and an open-source FaceTime SDK on GitHub. If you’ve built voice agents before, the main change is that the call arrives over FaceTime instead of a phone line.
Chert’s plans
Chert’s homepage lists three options.
| Free trial | Shared | Dedicated | |
|---|---|---|---|
| Price | $0 for 24 hours | $25/month | $250 per line per month |
| Line | Shared by up to 6 customers | Shared by up to 6 customers | Dedicated to your project |
| Users | 100 | 100 | Unlimited |
| FaceTime | 50 min a day per customer, 5 min per call | 50 min a day per customer, 5 min per call | Video needs a compatible agent and qualified line |
| Messages per hour | 50 | 100 per customer, 400 across the line | No rate limit |
| Agent can text first | No | No, only replies to users who joined | Yes |
| Inbound access | Verified phones or up to 100 allowlisted numbers | Verified phones or up to 100 allowlisted numbers | Open |
The structure makes Chert’s intended path clear. The trial and shared plans are for building and testing with known users. Anything customer-facing, where strangers text or call you or your agent starts the conversation, needs a dedicated line. Dedicated plans include business support.
How Getcho moved notifications to iMessage
Chert’s most detailed customer story is Getcho, a last-mile fulfillment platform that sends delivery updates to shoppers, drivers and store teams. In Chert’s telling, Getcho’s text notifications ran on SMS, which was slow when “your driver arrives in four minutes” mattered, read as spam to iPhone users, and handled group threads badly. Store customers started texting their human account reps on iMessage instead, because they preferred it.
Getcho tried RCS first. “In practice, most iPhones had RCS turned off and silently fell back to SMS,” according to the write-up. With Chert, the team set up iMessage group notification threads the same afternoon it got access, took about 15 minutes to onboard, and is moving every SMS notification over. This is Chert’s account, not an independent one, but the problems it describes with SMS and RCS on iPhones are ones many businesses will recognize.
Terms and data
Chert’s terms, last updated June 29, 2026, require customers to have consent from everyone they message and to comply with the TCPA and anti-spam laws. Two clauses stand out. Customers may not use Chert “for political campaigning, regulated industries, or other use cases prohibited by applicable platform or carrier rules” without Chert’s written consent, so healthcare and finance teams should ask first. And Chert can suspend access if it believes those rules have been broken.
We didn’t find a public security certification on Chert’s site. If you’ll send sensitive information, ask how message content is stored and who can access it.
How Chert compares
Each iMessage provider has settled on a different strength:
- Chert: FaceTime audio and video for agents, on top of iMessage and SMS.
- Linq: the infrastructure behind many consumer assistants, including Poke and Tomo, with managed load balancing across lines.
- Blooio: self-serve lines for small businesses and agencies, plus Apple Messages for Business.
- Photon: open source, and one codebase that also reaches WhatsApp and Telegram.
Who Chert suits
Chert is worth a look if:
- You’re building an agent people would rather talk to or see than text, such as a coach, tutor or companion.
- You already run voice agents on LiveKit and want FaceTime as another way in.
- You’re a business moving customer notifications or support off SMS and want help with the integration.
Look elsewhere if:
- You need a long operating history. Chert was founded in 2026.
- You work in a regulated industry and can’t wait for written approval.
- You need a published security certification before you start.
And if you’d rather have an AI that picks up your business phone and texts customers back today, without building an agent at all, that’s what Carly does. Add it to ChatGPT or Claude, get a business number, and its agents answer calls, text back missed callers and reply on WhatsApp, over regular phone lines and SMS rather than FaceTime and iMessage. Our guides to missed-call text back and AI text messaging show how that works.
FAQ
What is Chert?
Chert is an API from Chert Technologies, a Y Combinator Spring 2026 company, that gives AI agents and businesses a phone number for iMessage, SMS fallback and FaceTime audio and video calls.
Can an AI agent answer FaceTime calls with Chert?
Yes. Chert bridges FaceTime audio and video calls into a LiveKit room where your agent runs. Each line handles one active call at a time.
Who founded Chert?
Gary Gao (CEO) and Ian Fong (CTO). They launched Chert on Hacker News in May 2026.
Can Chert send the first message?
Only on a dedicated line. On the free trial and shared plan, the agent can reply to users who joined but can’t text anyone first.
Is Chert approved by Apple?
No. Chert runs iMessage and FaceTime outside Apple Messages for Business, Apple’s official business channel. Its founders say they only work with opt-in, non-spam use cases.
Does Chert fall back to SMS?
Yes. If a recipient isn’t on iMessage, Chert falls back to SMS and tells you through the API which channel the message used.
Ready to automate your busywork?
Carly schedules, researches, and briefs you—so you can focus on what matters.
See what people say
"Before Carly, I relied on a Calendly link, but the whole process felt impersonal and not very professional. Carly changed that by handling all the back-and-forth, so I'm no longer stuck in endless email threads trying to line up schedules.
Now Carly reaches out to candidates, shares my real-time availability, lets them pick a slot, then sends a Zoom link and drops it straight into my calendar. She sends reminders to both of us before each call, which has significantly reduced no-shows and last-minute confusion.
On top of scheduling, Carly acts like a full executive assistant, sending me my schedule the night before so I can prepare for each call. It reminds me of the old x.ai assistant, but Carly is noticeably smarter, faster, and better suited to my healthcare recruitment business."
