Text-First AI Agents Gain Momentum Across iMessage, SMS, and Messaging Apps

Other platforms have similarly attracted major backing or corporate exits. Tab emerged from stealth in October 2026 with a $300 million valuation backed by SV Angel, Valar Ventures, and American Spirit, allowing users on WhatsApp and iMessage to delegate errands like shopping and gift sourcing. In the workplace sector, Town raised a $55 million Series A round led by Andreessen Horowitz and Forerunner to build "Townie," an assistant that hooks into Slack, documents, calendars, and email to automate recurring office tasks.

Platform distribution is also shifting. In June 2026, Poke gained approval as the first AI agent on Apple Messages for Business, handling calendars, photo management, smart-home controls, and fitness. One month later, its parent entity, The Interaction Company of California, was bought by AI coding firm Cognition in a deal valued in the low nine figures.

Specialization across families, travel, and privacy

Beyond general assistants, developers are building text agents tailored to narrow operational contexts:

• Family logistics: Services like Fambot (which raised $3.5 million in pre-seed backing), Ohai, and Orbits (backed by a16z Speedrun, N49P, and Garage Capital) ingest school updates, sports schedules, and chore lists to generate automated evening briefings and shared calendars. Ollie addresses enterprise-grade household security by achieving SOC 2 compliance, offering tiered plans up to $100 per month.
• Autonomous identities and payments: Wajo launched its consumer agent, Fo, equipping the software with its own phone number, email address, and payment card so it can book appointments and purchase items without direct user credentials, falling back to human operators when automated paths fail.
• Travel and creators: Miso pairs iMessage interaction with human travel teams to track loyalty perks and manage flights. For creators, Stanley turns voice notes and Instagram history into social posts, scripts, and scheduled calendar entries.
• Local execution: While most agents rely on cloud servers—such as Folk, which runs on dedicated private cloud instances to execute code—Underdog operates locally on Mac and Windows devices to keep user information entirely off external servers.

Other tools focus on interface flexibility. Ask Proxy deploys across SMS for $1 per day to proactively offer reminders and gift suggestions, Caddy parses incoming texts and calendar invites on iMessage and RCS, and Comma provides an open-source, multi-device agent that checks its own work across browsers and files.

What it means for developers

For engineering teams, the emergence of messaging-native agents represents a departure from standard mobile and web frontend development. The primary interface is becoming a persistent text stream, which places the architectural burden entirely on conversational state management, webhook infrastructure, and tool invocation.

Building reliable agents over SMS or chat protocols requires orchestrating multiple backend components:

• Long-term memory and context: Agents such as Rene, Pally, and Martin must parse unstructured voice notes, forwarded emails, and images while maintaining state over weeks or months.
• Autonomous credentials: As seen with Instinct and Wajo, giving an agent its own email address, virtual card, or phone number requires strict boundary controls, permission gates (like Maverick's confirmation step), and human-in-the-loop fallback systems.
• Model diversity and routing: Different agent steps require distinct model strengths, from low-latency conversational parsing to heavy reasoning for multi-step task execution. Developers can try top AI models cheaply through one API at https://apixoai.online to evaluate how providers like Claude, GPT, Gemini, and DeepSeek handle complex tool calling and state tracking without managing separate vendor contracts.

As chat networks like iMessage, RCS, and WhatsApp become de facto runtime environments, developers will increasingly compete on reliability, third-party integrations, and autonomous execution rather than frontend UI design.

Source: Here are the top AI agents that can live in your text messages — TechCrunch AI. Written by the Apixo team from that report.